Patent Application
20240232590
Various embodiments of the present disclosure address technical challenges related to performing predictive data analysis and provide solutions to address the efficiency and reliability shortcomings of existing predictive data analysis solutions.
In general, various embodiments of the present disclosure provide methods, apparatus, systems, computing devices, computing entities, and/or the like for generating a representative embeddings for a plurality of temporal sequences by using a graph attention augmented temporal network based on dynamic co-occurrence graphs for preceding temporal sequences and initial embeddings, where the dynamic co-occurrence graphs are projections of a global guidance co-occurrence graph on classification features of the preceding temporal sequences, and the initial embeddings are generated by processing a latent representation of corresponding classification features that is generated by a sequential long short term memory model as well as a classification feature tree using a tree-based long short term memory model.
In accordance with one aspect, a method is provided. In one embodiment, the method comprises: receiving, by one or more processors, one or more input data objects, each input data object comprising (i) a temporal sequence in a plurality of temporal sequences, (ii) a related classification feature subset of a plurality of classification features associated with the temporal sequence, and (iii) a descriptive text feature associated with the temporal sequence; generating, by the one or more processors, a global guidance correlation graph data object, wherein: (i) each node of the global guidance correlation graph data object corresponds to a classification feature in the plurality of classification features, and (ii) each edge of the global guidance correlation graph data object corresponds to a classification feature pair and describes a co-occurrence probability for the classification feature pair; for each temporal sequence, generating, by the one or more processors, one or more dynamic co-occurrence graph data object based on the global guidance correlation graph, wherein each dynamic co-occurrence graph data object for a particular temporal sequence describes a projection of the global guidance correlation graph data object on the input data object for the temporal sequence; generating, by the one or more processors, using the machine learning model, and based on the plurality of temporal sequences and each dynamic co-occurrence graph data object, a plurality of prediction classification features, wherein: the machine learning model comprises a graph-attention augmented temporal neural network machine learning model comprising a plurality of embedding layers, training the machine learning model comprises, for each combination of a given temporal sequence t of T number of temporal sequences in the plurality of temporal sequences, a given non-initial embedding layer l of the plurality embedding layers, and a given classification feature i of the plurality of classification features, a) generating a historical node representation based on: (i) a prior-layer historical node representation for the given temporal sequence t and the given classification feature i as generated by a preceding embedding layer l−1, and (ii) neighbor nodes for a target node associated with the given classification feature i in the dynamic co-occurrence graph corresponding to the given temporal sequence t, and b) appending an attention vector comprising a node attention layer associated with the descriptive text feature to the historical node representation, an initial embedding layer is configured to, for an initial temporal sequence, generate historical node representations for the plurality of classification features using a tree-of-sequences based on initial embeddings that are generated using a sequential long short-term memory machine learning model; generating, by the one or more processors, one or more predicted edges in the global guidance correlation graph data object, wherein the one or more predicted edges are connected to at least one node of the global guidance correlation graph data object associated with the plurality of prediction classification features; determining, by the one or more processors, one or more lowest common ancestor nodes of the plurality of prediction classification features from a classification feature graph based on the one or more predicted edges; generating, by the one or more processors, one or more link predictions based on the one or more lowest common ancestor nodes; and initiating performance, by the one or more processors, of one or more prediction-based actions based on the one or more link predictions.
In accordance with another aspect, an apparatus comprising at least one processor and at least one memory including computer program code is provided. In one embodiment, the at least one memory and the computer program code may be configured to, with the processor, cause the apparatus to: receive one or more input data objects, each input data object comprising (i) a temporal sequence in a plurality of temporal sequences, (ii) a related classification feature subset of a plurality of classification features associated with the temporal sequence, and (iii) a descriptive text feature associated with the temporal sequence; generate a global guidance correlation graph data object, wherein: (i) each node of the global guidance correlation graph data object corresponds to a classification feature in the plurality of classification features, and (ii) each edge of the global guidance correlation graph data object corresponds to a classification feature pair and describes a co-occurrence probability for the classification feature pair; for each temporal sequence, generate one or more dynamic co-occurrence graph data object based on the global guidance correlation graph, wherein each dynamic co-occurrence graph data object for a particular temporal sequence describes a projection of the global guidance correlation graph data object on the input data object for the temporal sequence; generate, using the machine learning model and based on the plurality of temporal sequences and each dynamic co-occurrence graph data object, a plurality of prediction classification features, wherein: the machine learning model comprises a graph-attention augmented temporal neural network machine learning model comprising a plurality of embedding layers, training the machine learning model comprises, for each combination of a given temporal sequence t of T number of temporal sequences in the plurality of temporal sequences, a given non-initial embedding layer l of the plurality embedding layers, and a given classification feature i of the plurality of classification features, a) generating a historical node representation based on: (i) a prior-layer historical node representation for the given temporal sequence t and the given classification feature i as generated by a preceding embedding layer l−1, and (ii) neighbor nodes for a target node associated with the given classification feature i in the dynamic co-occurrence graph corresponding to the given temporal sequence t, and b) appending an attention vector comprising a node attention layer associated with the descriptive text feature to the historical node representation, an initial embedding layer is configured to, for an initial temporal sequence, generate historical node representations for the plurality of classification features using a tree-of-sequences based on initial embeddings that are generated using a sequential long short-term memory machine learning model; generate one or more predicted edges in the global guidance correlation graph data object, wherein the one or more predicted edges are connected to at least one node of the global guidance correlation graph data object associated with the plurality of prediction classification features; determine one or more lowest common ancestor nodes of the plurality of prediction classification features from a classification feature graph based on the one or more predicted edges; generate one or more link predictions based on the one or more lowest common ancestor nodes; and initiate performance of one or more prediction-based actions based on the one or more link predictions.
In accordance with yet another aspect, a computer program product is provided. The computer program product may comprise at least one computer-readable storage medium having computer-readable program code portions stored therein, the computer-readable program code portions comprising executable portions configured to: receive one or more input data objects, each input data object comprising (i) a temporal sequence in a plurality of temporal sequences, (ii) a related classification feature subset of a plurality of classification features associated with the temporal sequence, and (iii) a descriptive text feature associated with the temporal sequence; generate a global guidance correlation graph data object, wherein: (i) each node of the global guidance correlation graph data object corresponds to a classification feature in the plurality of classification features, and (ii) each edge of the global guidance correlation graph data object corresponds to a classification feature pair and describes a co-occurrence probability for the classification feature pair; for each temporal sequence, generate one or more dynamic co-occurrence graph data object based on the global guidance correlation graph, wherein each dynamic co-occurrence graph data object for a particular temporal sequence describes a projection of the global guidance correlation graph data object on the input data object for the temporal sequence; generate, using the machine learning model and based on the plurality of temporal sequences and each dynamic co-occurrence graph data object, a plurality of prediction classification features, wherein: the machine learning model comprises a graph-attention augmented temporal neural network machine learning model comprising a plurality of embedding layers, training the machine learning model comprises, for each combination of a given temporal sequence t of T number of temporal sequences in the plurality of temporal sequences, a given non-initial embedding layer l of the plurality embedding layers, and a given classification feature i of the plurality of classification features, a) generating a historical node representation based on: (i) a prior-layer historical node representation for the given temporal sequence t and the given classification feature i as generated by a preceding embedding layer l−1, and (ii) neighbor nodes for a target node associated with the given classification feature i in the dynamic co-occurrence graph corresponding to the given temporal sequence t, and b) appending an attention vector comprising a node attention layer associated with the descriptive text feature to the historical node representation, an initial embedding layer is configured to, for an initial temporal sequence, generate historical node representations for the plurality of classification features using a tree-of-sequences based on initial embeddings that are generated using a sequential long short-term memory machine learning model; generate one or more predicted edges in the global guidance correlation graph data object, wherein the one or more predicted edges are connected to at least one node of the global guidance correlation graph data object associated with the plurality of prediction classification features; determine one or more lowest common ancestor nodes of the plurality of prediction classification features from a classification feature graph based on the one or more predicted edges; generate one or more link predictions based on the one or more lowest common ancestor nodes; and initiate performance of one or more prediction-based actions based on the one or more link predictions.
Having thus described the present disclosure in general terms, reference will now be made to the accompanying drawings, which are not necessarily drawn to scale, and wherein:
Various embodiments of the present disclosure are described more fully hereinafter with reference to the accompanying drawings, in which some, but not all embodiments of the present disclosure are shown. Indeed, the present disclosure may be embodied in many different forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided so that this present disclosure will satisfy applicable legal requirements. The term “or” is used herein in both the alternative and conjunctive sense, unless otherwise indicated. The terms “illustrative” and “example” are used to be examples with no indication of quality level. Terms such as “computing,” “determining,” “generating,” and/or similar words are used herein interchangeably to refer to the creation, modification, or identification of data. Further, “based on,” “based at least in part on,” “based at least on,” “based upon,” and/or similar words are used herein interchangeably in an open-ended manner such that they do not indicate being based only on or based solely on the referenced element or elements unless so indicated. Like numbers refer to like elements throughout.
Various embodiments of the present disclosure make important technical contributions to improving predictive accuracy of predictive machine learning models, which in turn improves training speed and training efficiency of machine learning models. This approach improves training speed and training efficiency of training predictive machine learning models. It is well-understood in the relevant art that there is typically a tradeoff between predictive accuracy and training speed, such that it is trivial to improve training speed by reducing predictive accuracy. Thus the real challenge is to improve training speed without sacrificing predictive accuracy through innovative model architectures. Accordingly, techniques that improve predictive accuracy without harming training speed, such as the techniques described herein, enable improving training speed given a constant predictive accuracy. In doing so, the techniques described herein improve efficiency and speed of training predictive machine learning models, thus reducing the number of computational operations needed and/or the amount of training data entries needed to train predictive machine learning models. Accordingly, the techniques described herein improve at least one of the computational efficiency, storage-wise efficiency, and speed of training machine learning models.
For example, various embodiments of the present disclosure improve accuracy of machine learning models by pre-training models with classification features (e.g., clinical events) based on textual content as well as hierarchical structure among the classification features. As described herein, a collection of information, such as an electronic health record (EHR), may comprise a large number of classification features associated with a temporal structure. Existing methods for performing a prediction on a collection of information may include building a prediction model based on the classification features and making a prediction. However, existing methods are limited in their abilities in dealing with complex structural correlations and temporal dependencies of classification features in a temporal sequence (e.g., admission), which may impact classification predictions as well as temporal prediction of classification features.
However, in accordance with various embodiments of the present disclosure, a temporal-spatial approach may be used to capture temporal classification feature set (e.g., health) progression as well as relationships among different classification features over a period. This technique will lead to higher accuracy of performing predictions on input comprising a temporal sequence. In doing so, the techniques described herein improving efficiency and speed of training natural language processing machine learning models, thus reducing the number of computational operations needed and/or the amount of training data entries needed to train machine learning models. Accordingly, the techniques described herein improve at least one of the computational efficiency, storage-wise efficiency, and speed of training machine learning models.
An example application of various embodiments of the present disclosure relates to generating predictions for a given temporal sequence based on classification features from a set of temporal sequences. In some embodiments, an attentional encoder decoder model may be trained on a collection of information comprising a set of temporal sequences, where weights from the attentional encoder decoder model are transferred to a graph-attention augmented temporal neural network model which results in a better parameter initialization of the graph attention model. A key benefit of various embodiments of the present disclosure is the improved prediction outcome based on textual content as well as any hierarchical structure among text within an EHR. This improved accuracy of prediction also enables improved accuracy in further text processing tasks, such as coding quality and diagnosis. In some embodiments, the following operations are performed: generating a global guidance graph where each node is a diagnostic event; generating a dynamic co-occurrence graph for each temporal sequence weighted by co-occurrences of events; initializing embeddings using pretrained models and using temporal sequence level representation to learn temporal classification feature set progression using a graph attention augmented temporal neural network; and generating a prediction for a future temporal sequence.
In some embodiments, recommendation of a current admission may be generated by considering a patient's historical records and correlations among clinical events from every admission in the patient's historical records. Accordingly, various embodiments of the present disclosure deal with complex structural correlations and temporal dependencies of clinical events in EHRs, which results in improved recommendation quality and temporal prediction ability in the context of providing healthcare.
The term “temporal sequence” may refer to a data construct that describes an input data object comprising an instance xt={dt,et} at a given instance time t, for a given entity, where dt is representative of one or more classification features that describe the instance xt, and et is representative of descriptive text feature associated with the instance xt. A temporal sequence may be descriptive of one or more events or actions associated with an entity for instance xt. For example, the instance may represent a patient admission on a given date and the given entity may represent a patient. As such, a temporal sequence may be representative of one or more actions for the instance xt. A temporal sequence may comprise images, text files, audio/video files, and application files that may be used to, for example, train a machine learning model. Classification features within a temporal sequence may be internally related and include correlations of various degrees that may be interpreted with various meanings.
The term “set of temporal sequences” may refer to a data construct that describes a plurality of temporal sequences En={x1n, x2n, . . . , x{T(n)}n} of a given entity n for T(n) number of instances xt for instance times 1-t. In some embodiments, the plurality of temporal sequences may be an ordered set of temporal sequences. As an example, the set of temporal sequences may comprise a patient's EHRs comprising a history of admissions where each of the plurality of temporal sequences may represent an admission within a time series. As such, a set of temporal sequences may be representative of a patient's condition over multiple admissions, procedures, and medications.
The term “classification feature” may refer to a data construct that describes an attribute or characteristic associated with an input data object, such as a temporal sequence, which may be used for analysis and training of a machine learning model. A classification feature may comprise descriptions, tags, or identifiers that classify or emphasize characteristics present in a body of data. As an example, classification features of an input data object may comprise clinical events including diagnoses, associated symptoms, or medical codes, e.g., International Statistical Classification of Diseases and Related Health Problems (ICD) codes, Current Procedural Terminology (CPT) codes, prescription (RX) codes.
The term “descriptive text feature” may refer to a data construct that describes a textual description comprising a narrative description associated with a temporal sequence. For example, a descriptive text feature may comprise text describing one or more procedures performed on a patient.
The term “initial embedding” may refer to a data construct that describes an initial representation of classification features from input data objects, where a classification feature matrix may be generated by processing the input data objects using an attentional encoder decoder machine learning model. For example, the initial embedding may comprise a translation of classification features from a temporal sequence to a classification feature matrix comprising a tree hierarchy using an input embedding module of the attentional encoder decoder machine learning model.
The term “attentional encoder decoder machine learning model” may refer to a data construct that describes parameters, hyperparameters, and/or defined operations of a machine learning model, where the machine learning model is configured to translate classification features into a classification feature matrix by processing input data objects. The attentional encoder decoder machine learning model may comprise at least an input embedding module that translates classification features from a temporal sequence to a classification feature matrix. In some embodiments, the classification feature matrix may be created by using a tree-of-sequences LSTM network. According to various embodiments of the present disclosure, the input embedding module may include a description encoder that generates a latent representation vector for a description corresponding to a classification feature. In some embodiments, each classification feature may be associated with a description that describes the semantics of the classification features. For example, a classification feature, such as a diagnostic code, may be associated with a short text semantic description that describes the semantics of the diagnostic code. Other examples of such a description may include metadata or captioning. The description encoder may use a sequential long short-term memory (LSTM) network to encode the short text semantic descriptions. The input embedding module may further include a classification feature encoder that creates a tree-of-sequences LSTM network to capture hierarchal relationships among the classification features. Each node of the tree-of-sequences LSTM network may comprise an input vector based on the latent representation generated by the sequential LSTM.
The term “LSTM network” may refer to a data construct that describes a recurrent neural network that stores a representation of classification features from sequences of input data objects, such as a set of temporal sequences, where long- and short-term information dependencies are preserved.
The term “sequential LSTM network” may refer to a data construct that describes a LSTM network that learns a latent representation of classification features (which may reflect certain semantic information) and models a sequential structure among the classification features.
The term “tree-of-sequences LSTM network” may refer to a data construct that describes a hierarchy of sequential LSTM networks. As described above, the tree-of-sequences LSTM network comprises classification features that are encoded by a sequential LSTM and applying a tree structure to the sequential LSTM to capture hierarchical relationships among the classification features. Each node of the tree-of-sequences LSTM network may comprise a vector including a latent representation based on the encoding by the sequential LSTM.
The term “global guidance correlation graph” may refer to a data construct that describes a graph including nodes that are representative of a universe of classification features appearing in a data set (e.g., of input data objects). For example, the nodes of a global guidance correlation graph may include all classification features in a dataset comprising a plurality of sets of temporal sequences associated with a plurality of entities. The global guidance correlation graph may further include edges that are based on co-occurrence probability between the universe of classification features. As an example, edges of a global guidance correlation graph may include weights that may be calculated based on total number of temporal sequences that a classification feature pair have co-occurred, total number of temporal sequences that classification features of the pair have appeared at least once, and total number of temporal sequences.
The term “dynamic co-occurrence graphs” may refer to a data construct that describes a plurality of semantic graphs including classification features associated with a plurality of temporal sequences over time. A dynamic co-occurrence graph may be constructed based on global correlations from a global guidance correlation graph. In some embodiments, dynamic co-occurrence graphs may comprise a sequence of adjacency matrices where each adjacency matrix in the sequence of adjacency matrices comprises a connected graph including nodes that represent classification features of a given temporal sequence of a set of temporal sequences associated with a given entity. The adjacency matrices may further include edges and edge weights based on a global guidance correlation graph.
The term “co-occurrence probability” may refer to a data construct that describes a measurement of relationship between two variables. In some embodiments, the relationship may comprise a semantic proximity of the two variables. For example, co-occurrence frequency may comprise an above-chance frequency of classification features coinciding or existing within a body of data.
The term “graph-attention augmented temporal neural network” may refer to a data construct that describes a two-layer graph-attention neural network for embedding historical node representations. In some embodiments, the graph-attention augmented temporal neural network model may update an initial embedding based on a dynamic co-occurrence graph. The graph-attention augmented temporal neural network may construct encoded classification feature vectors that contain information of other co-occurrence classification features of a same temporal sequence with different degrees of correlation to obtain more comprehensive representation. As an example, at each embedding layer, the graph-attention augmented temporal neural network can embed a set of historical node representations by recursively aggregating information from node neighbors based on the dynamic co-occurrence graphs.
The term “prediction classification feature” may refer to a data construct that describes an output of a machine learning model, such as a graph-attention augmented temporal neural network model, based on a given input data object. As an example, in some embodiments, prediction classification features may comprise medical codes, e.g., ICD codes, CPT codes, and RX codes that are generated for prediction on a set of temporal sequences. According to various embodiments of the present disclosure, a prediction classification feature may comprise a diagnostic code for a class of disease yt={0,1}L+1 at instance time t, where L is the number of classes of diseases considered, and all other diseases are represented as a single binary vector.
The term “link prediction” may refer to a data construct that describes a predicted classification of classification features identified from a classification feature graph based on one or more predicted edges. A link prediction may comprise an identification of a classification feature associated with a lowest common ancestor node.
The term “predicted edge” may refer to a data construct that describes a likely edge between two nodes of a global guidance correlation graph data object. The likely edge may be determined based on a vector generated by concatenating at least two nodes selected from the global guidance correlation graph data object. At least one of the two nodes may be associated with prediction classification features.
The term “lowest common ancestor node” may refer to a data construct that describes a root node of a smallest subtree within a classification feature graph including child nodes associated with at least one predicted edge within a global guidance correlation graph data object.
Embodiments of the present disclosure may be implemented in various ways, including as computer program products that comprise articles of manufacture. Such computer program products may include one or more software components including, for example, software objects, methods, data structures, or the like. A software component may be coded in any of a variety of programming languages. An illustrative programming language may be a lower-level programming language, such as an assembly language, associated with a particular hardware architecture and/or operating system platform. A software component comprising assembly language instructions may require conversion into executable machine code by an assembler prior to execution by the hardware architecture and/or platform. Another example programming language may be a higher-level programming language that may be portable across multiple architectures. A software component comprising higher-level programming language instructions may require conversion to an intermediate representation by an interpreter or a compiler prior to execution.
Other examples of programming languages include, but are not limited to, a macro language, a shell or command language, a job control language, a script language, a database query or search language, and/or a report writing language. In one or more example embodiments, a software component comprising instructions in one of the foregoing examples of programming languages may be executed directly by an operating system or other software component without having to be first transformed into another form. A software component may be stored as a file or other data storage construct. Software components of a similar type or functionally related may be stored together, such as in a particular directory, folder, or library. Software components may be static (e.g., pre-established or fixed) or dynamic (e.g., created or modified at the time of execution).
A computer program product may include a non-transitory computer-readable storage medium storing applications, programs, program modules, scripts, source code, program code, object code, byte code, compiled code, interpreted code, machine code, executable instructions, and/or the like (also referred to herein as executable instructions, instructions for execution, computer program products, program code, and/or similar terms used herein interchangeably). Such non-transitory computer-readable storage media include all computer-readable media (including volatile and non-volatile media).
In one embodiment, a non-volatile computer-readable storage medium may include a floppy disk, flexible disk, hard disk, solid-state storage (SSS) (e.g., a solid state drive (SSD), solid state card (SSC), solid state module (SSM), enterprise flash drive, magnetic tape, or any other non-transitory magnetic medium, and/or the like. A non-volatile computer-readable storage medium may also include a punch card, paper tape, optical mark sheet (or any other physical medium with patterns of holes or other optically recognizable indicia), compact disc read only memory (CD-ROM), compact disc-rewritable (CD-RW), digital versatile disc (DVD), Blu-ray disc (BD), any other non-transitory optical medium, and/or the like. Such a non-volatile computer-readable storage medium may also include read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), flash memory (e.g., Serial, NAND, NOR, and/or the like), multimedia memory cards (MMC), secure digital (SD) memory cards, SmartMedia cards, CompactFlash (CF) cards, Memory Sticks, and/or the like. Further, a non-volatile computer-readable storage medium may also include conductive-bridging random access memory (CBRAM), phase-change random access memory (PRAM), ferroelectric random-access memory (FeRAM), non-volatile random-access memory (NVRAM), magnetoresistive random-access memory (MRAM), resistive random-access memory (RRAM), Silicon-Oxide-Nitride-Oxide-Silicon memory (SONOS), floating junction gate random access memory (FJG RAM), Millipede memory, racetrack memory, and/or the like.
In one embodiment, a volatile computer-readable storage medium may include random access memory (RAM), dynamic random access memory (DRAM), static random access memory (SRAM), fast page mode dynamic random access memory (FPM DRAM), extended data-out dynamic random access memory (EDO DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDR SDRAM), double data rate type two synchronous dynamic random access memory (DDR2 SDRAM), double data rate type three synchronous dynamic random access memory (DDR3 SDRAM), Rambus dynamic random access memory (RDRAM), Twin Transistor RAM (TTRAM), Thyristor RAM (T-RAM), Zero-capacitor (Z-RAM), Rambus in-line memory module (RIMM), dual in-line memory module (DIMM), single in-line memory module (SIMM), video random access memory (VRAM), cache memory (including various levels), flash memory, register memory, and/or the like. It will be appreciated that where embodiments are described to use a computer-readable storage medium, other types of computer-readable storage media may be substituted for or used in addition to the computer-readable storage media described above.
As should be appreciated, various embodiments of the present disclosure may also be implemented as methods, apparatus, systems, computing devices, computing entities, and/or the like. As such, embodiments of the present disclosure may take the form of an apparatus, system, computing device, computing entity, and/or the like executing instructions stored on a computer-readable storage medium to perform certain steps or operations. Thus, embodiments of the present disclosure may also take the form of an entirely hardware embodiment, an entirely computer program product embodiment, and/or an embodiment that comprises combination of computer program products and hardware performing certain steps or operations. Embodiments of the present disclosure are described below with reference to block diagrams and flowchart illustrations. Thus, it should be understood that each block of the block diagrams and flowchart illustrations may be implemented in the form of a computer program product, an entirely hardware embodiment, a combination of hardware and computer program products, and/or apparatus, systems, computing devices, computing entities, and/or the like carrying out instructions, operations, steps, and similar words used interchangeably (e.g., the executable instructions, instructions for execution, program code, and/or the like) on a computer-readable storage medium for execution. For example, retrieval, loading, and execution of code may be performed sequentially such that one instruction is retrieved, loaded, and executed at a time. In some example embodiments, retrieval, loading, and/or execution may be performed in parallel such that multiple instructions are retrieved, loaded, and/or executed together. Thus, such embodiments can produce specifically-configured machines performing the steps or operations specified in the block diagrams and flowchart illustrations. Accordingly, the block diagrams and flowchart illustrations support various combinations of embodiments for performing the specified instructions, operations, or steps.
An example of a prediction-based action that can be performed using the predictive data analysis system 101 comprises receiving a request for generating a diagnostic code for a class of disease based on an EHR of a patient and predicting associated co-morbidities with the predicted disease class, and displaying the diagnostic code and predicted co-morbidities on a user interface. Other examples of prediction-based actions comprise generating a diagnostic report, displaying/providing resources, generating action scripts, generating alerts or reminders, and generating one or more electronic communications based on the predicted disease class.
In accordance with various embodiments of the present disclosure, a graph-attention augmented temporal neural network model may be used to generate prediction classification features on a set of temporal sequences based on temporal classification feature set progression as well as relationships among different classification features over a period. This technique will lead to higher accuracy of performing predictions on input comprising a temporal sequence. In doing so, the techniques described herein improving efficiency and speed of training natural language processing machine learning models, thus reducing the number of computational operations needed and/or the amount of training data entries needed to train machine learning models. Accordingly, the techniques described herein improve at least one of the computational efficiency, storage-wise efficiency, and speed of training machine learning models.
In some embodiments, predictive data analysis system 101 may communicate with at least one of the client computing entities 102 using one or more communication networks. Examples of communication networks include any wired or wireless communication network including, for example, a wired or wireless local area network (LAN), personal area network (PAN), metropolitan area network (MAN), wide area network (WAN), or the like, as well as any hardware, software and/or firmware required to implement it (such as, e.g., network routers, and/or the like).
The predictive data analysis system 101 may include a predictive data analysis computing entity 106 and a storage subsystem 108. The predictive data analysis computing entity 106 may be configured to receive predictive data analysis requests from one or more client computing entities 102, process the predictive data analysis requests to generate predictions corresponding to the predictive data analysis requests, provide the generated predictions to the client computing entities 102, and automatically perform prediction-based actions based on the generated predictions.
The storage subsystem 108 may be configured to store input data used by the predictive data analysis computing entity 106 to perform predictive data analysis as well as model definition data used by the predictive data analysis computing entity 106 to perform various predictive data analysis tasks. The storage subsystem 108 may include one or more storage units, such as multiple distributed storage units that are connected through a computer network. Each storage unit in the storage subsystem 108 may store at least one of one or more data assets and/or one or more data about the computed properties of one or more data assets. Moreover, each storage unit in the storage subsystem 108 may include one or more non-volatile storage or memory media including, but not limited to, hard disks, ROM, PROM, EPROM, EEPROM, flash memory, MMCs, SD memory cards, Memory Sticks, CBRAM, PRAM, FeRAM, NVRAM, MRAM, RRAM, SONOS, FJG RAM, Millipede memory, racetrack memory, and/or the like.
As indicated, in one embodiment, the predictive data analysis computing entity 106 may also include one or more communications interfaces 220 for communicating with various computing entities, such as by communicating data, content, information, and/or similar terms used herein interchangeably that can be transmitted, received, operated on, processed, displayed, stored, and/or the like.
As shown in
For example, the processing element 205 may be embodied as one or more complex programmable logic devices (CPLDs), microprocessors, multi-core processors, coprocessing entities, application-specific instruction-set processors (ASIPs), microcontrollers, and/or controllers. Further, the processing element 205 may be embodied as one or more other processing devices or circuitry. The term circuitry may refer to an entirely hardware embodiment or a combination of hardware and computer program products. Thus, the processing element 205 may be embodied as integrated circuits, application specific integrated circuits (ASICs), field programmable gate arrays (FPGAs), programmable logic arrays (PLAs), hardware accelerators, other circuitry, and/or the like.
As will therefore be understood, the processing element 205 may be configured for a particular use or configured to execute instructions stored in volatile or non-volatile media or otherwise accessible to the processing element 205. As such, whether configured by hardware or computer program products, or by a combination thereof, the processing element 205 may be capable of performing steps or operations according to embodiments of the present disclosure when configured accordingly.
In one embodiment, the predictive data analysis computing entity 106 may further include, or be in communication with, non-volatile media (also referred to as non-volatile storage, memory, memory storage, memory circuitry and/or similar terms used herein interchangeably). In one embodiment, the non-volatile storage or memory may include one or more non-volatile storage or memory media 210, including, but not limited to, hard disks, ROM, PROM, EPROM, EEPROM, flash memory, MMCs, SD memory cards, Memory Sticks, CBRAM, PRAM, FeRAM, NVRAM, MRAM, RRAM, SONOS, FJG RAM, Millipede memory, racetrack memory, and/or the like.
As will be recognized, the non-volatile storage or memory media may store databases, database instances, database management systems, data, applications, programs, program modules, scripts, source code, object code, byte code, compiled code, interpreted code, machine code, executable instructions, and/or the like. The term database, database instance, database management system, and/or similar terms used herein interchangeably may refer to a collection of records or data that is stored in a computer-readable storage medium using one or more database models, such as a hierarchical database model, network model, relational model, entity-relationship model, object model, document model, semantic model, graph model, and/or the like.
In one embodiment, the predictive data analysis computing entity 106 may further include, or be in communication with, volatile media (also referred to as volatile storage, memory, memory storage, memory circuitry and/or similar terms used herein interchangeably). In one embodiment, the volatile storage or memory may also include one or more volatile storage or memory media 215, including, but not limited to, RAM, DRAM, SRAM, FPM DRAM, EDO DRAM, SDRAM, DDR SDRAM, DDR2 SDRAM, DDR3 SDRAM, RDRAM, TTRAM, T-RAM, Z-RAM, RIMM, DIMM, SIMM, VRAM, cache memory, register memory, and/or the like.
As will be recognized, the volatile storage or memory media may be used to store at least portions of the databases, database instances, database management systems, data, applications, programs, program modules, scripts, source code, object code, byte code, compiled code, interpreted code, machine code, executable instructions, and/or the like being executed by, for example, the processing element 205. Thus, the databases, database instances, database management systems, data, applications, programs, program modules, scripts, source code, object code, byte code, compiled code, interpreted code, machine code, executable instructions, and/or the like may be used to control certain aspects of the operation of the predictive data analysis computing entity 106 with the assistance of the processing element 205 and operating system.
As indicated, in one embodiment, the predictive data analysis computing entity 106 may also include one or more communications interfaces 220 for communicating with various computing entities, such as by communicating data, content, information, and/or similar terms used herein interchangeably that can be transmitted, received, operated on, processed, displayed, stored, and/or the like. Such communication may be executed using a wired data transmission protocol, such as fiber distributed data interface (FDDI), digital subscriber line (DSL), Ethernet, asynchronous transfer mode (ATM), frame relay, data over cable service interface specification (DOCSIS), or any other wired transmission protocol. Similarly, the predictive data analysis computing entity 106 may be configured to communicate via wireless external communication networks using any of a variety of protocols, such as general packet radio service (GPRS), Universal Mobile Telecommunications System (UMTS), Code Division Multiple Access 2000 (CDMA2000), CDMA2000 1× (1×RTT), Wideband Code Division Multiple Access (WCDMA), Global System for Mobile Communications (GSM), Enhanced Data rates for GSM Evolution (EDGE), Time Division-Synchronous Code Division Multiple Access (TD-SCDMA), Long Term Evolution (LTE), Evolved Universal Terrestrial Radio Access Network (E-UTRAN), Evolution-Data Optimized (EVDO), High Speed Packet Access (HSPA), High-Speed Downlink Packet Access (HSDPA), IEEE 802.11 (Wi-Fi), Wi-Fi Direct, 802.16 (WiMAX), ultra-wideband (UWB), infrared (IR) protocols, near field communication (NFC) protocols, Wibree, Bluetooth protocols, wireless universal serial bus (USB) protocols, and/or any other wireless protocol.
Although not shown, the predictive data analysis computing entity 106 may include, or be in communication with, one or more input elements, such as a keyboard input, a mouse input, a touch screen/display input, motion input, movement input, audio input, pointing device input, joystick input, keypad input, and/or the like. The predictive data analysis computing entity 106 may also include, or be in communication with, one or more output elements (not shown), such as audio output, video output, screen/display output, motion output, movement output, and/or the like.
The signals provided to and received from the transmitter 304 and the receiver 306, correspondingly, may include signaling information/data in accordance with air interface standards of applicable wireless systems. In this regard, the client computing entity 102 may be capable of operating with one or more air interface standards, communication protocols, modulation types, and access types. More particularly, the client computing entity 102 may operate in accordance with any of a number of wireless communication standards and protocols, such as those described above with regard to the predictive data analysis computing entity 106. In a particular embodiment, the client computing entity 102 may operate in accordance with multiple wireless communication standards and protocols, such as UMTS, CDMA2000, 1×RTT, WCDMA, GSM, EDGE, TD-SCDMA, LTE, E-UTRAN, EVDO, HSPA, HSDPA, Wi-Fi, Wi-Fi Direct, WiMAX, UWB, IR, NFC, Bluetooth, USB, and/or the like. Similarly, the client computing entity 102 may operate in accordance with multiple wired communication standards and protocols, such as those described above with regard to the predictive data analysis computing entity 106 via a network interface 320.
Via these communication standards and protocols, the client computing entity 102 can communicate with various other entities using concepts such as Unstructured Supplementary Service Data (USSD), Short Message Service (SMS), Multimedia Messaging Service (MMS), Dual-Tone Multi-Frequency Signaling (DTMF), and/or Subscriber Identity Module Dialer (SIM dialer). The client computing entity 102 can also download changes, add-ons, and updates, for instance, to its firmware, software (e.g., including executable instructions, applications, program modules), and operating system.
According to one embodiment, the client computing entity 102 may include location determining aspects, devices, modules, functionalities, and/or similar words used herein interchangeably. For example, the client computing entity 102 may include outdoor positioning aspects, such as a location module adapted to acquire, for example, latitude, longitude, altitude, geocode, course, direction, heading, speed, universal time (UTC), date, and/or various other information/data. In one embodiment, the location module can acquire data, sometimes known as ephemeris data, by identifying the number of satellites in view and the relative positions of those satellites (e.g., using global positioning systems (GPS)). The satellites may be a variety of different satellites, including Low Earth Orbit (LEO) satellite systems, Department of Defense (DOD) satellite systems, the European Union Galileo positioning systems, the Chinese Compass navigation systems, Indian Regional Navigational satellite systems, and/or the like. This data can be collected using a variety of coordinate systems, such as the Decimal Degrees (DD); Degrees, Minutes, Seconds (DMS); Universal Transverse Mercator (UTM); Universal Polar Stereographic (UPS) coordinate systems; and/or the like. Alternatively, the location information/data can be determined by triangulating the client computing entity's 102 position in connection with a variety of other systems, including cellular towers, Wi-Fi access points, and/or the like. Similarly, the client computing entity 102 may include indoor positioning aspects, such as a location module adapted to acquire, for example, latitude, longitude, altitude, geocode, course, direction, heading, speed, time, date, and/or various other information/data. Some of the indoor systems may use various position or location technologies including RFID tags, indoor beacons or transmitters, Wi-Fi access points, cellular towers, nearby computing devices (e.g., smartphones, laptops) and/or the like. For instance, such technologies may include the iBeacons, Gimbal proximity beacons, Bluetooth Low Energy (BLE) transmitters, NFC transmitters, and/or the like. These indoor positioning aspects can be used in a variety of settings to determine the location of someone or something to within inches or centimeters.
The client computing entity 102 may also comprise a user interface (that can include a display 316 coupled to a processing element 308) and/or a user input interface (coupled to a processing element 308). For example, the user interface may be a user application, browser, user interface, and/or similar words used herein interchangeably executing on and/or accessible via the client computing entity 102 to interact with and/or cause display of information/data from the predictive data analysis computing entity 106, as described herein. The user input interface can comprise any of a number of devices or interfaces allowing the client computing entity 102 to receive data, such as a keypad 318 (hard or soft), a touch display, voice/speech or motion interfaces, or other input device. In embodiments including a keypad 318, the keypad 318 can include (or cause display of) the conventional numeric (0-9) and related keys (#, *), and other keys used for operating the client computing entity 102 and may include a full set of alphabetic keys or set of keys that may be activated to provide a full set of alphanumeric keys. In addition to providing input, the user input interface can be used, for example, to activate or deactivate certain functions, such as screen savers and/or sleep modes.
The client computing entity 102 can also include volatile storage or memory 322 and/or non-volatile storage or memory 324, which can be embedded and/or may be removable. For example, the non-volatile memory may be ROM, PROM, EPROM, EEPROM, flash memory, MMCs, SD memory cards, Memory Sticks, CBRAM, PRAM, FeRAM, NVRAM, MRAM, RRAM, SONOS, FJG RAM, Millipede memory, racetrack memory, and/or the like. The volatile memory may be RAM, DRAM, SRAM, FPM DRAM, EDO DRAM, SDRAM, DDR SDRAM, DDR2 SDRAM, DDR3 SDRAM, RDRAM, TTRAM, T-RAM, Z-RAM, RIMM, DIMM, SIMM, VRAM, cache memory, register memory, and/or the like. The volatile and non-volatile storage or memory can store databases, database instances, database management systems, data, applications, programs, program modules, scripts, source code, object code, byte code, compiled code, interpreted code, machine code, executable instructions, and/or the like to implement the functions of the client computing entity 102. As indicated, this may include a user application that is resident on the entity or accessible through a browser or other user interface for communicating with the predictive data analysis computing entity 106 and/or various other computing entities.
In another embodiment, the client computing entity 102 may include one or more components or functionality that are the same or similar to those of the predictive data analysis computing entity 106, as described in greater detail above. As will be recognized, these architectures and descriptions are provided for example purposes only and are not limiting to the various embodiments.
In various embodiments, the client computing entity 102 may be embodied as an artificial intelligence (AI) computing entity, such as an Amazon Echo, Amazon Echo Dot, Amazon Show, Google Home, and/or the like. Accordingly, the client computing entity 102 may be configured to provide and/or receive information/data from a user via an input/output mechanism, such as a display, a camera, a speaker, a voice-activated input, and/or the like. In certain embodiments, an AI computing entity may comprise one or more predefined and executable program algorithms stored within an onboard memory storage module, and/or accessible over a network. In various embodiments, the AI computing entity may be configured to retrieve and/or execute one or more of the predefined program algorithms upon the occurrence of a predefined trigger event.
As described below, various embodiments of the present disclosure make important technical contributions to improving predictive accuracy of predictive machine learning models, which in turn improves training speed and training efficiency of machine learning models. This approach improves training speed and training efficiency of training predictive machine learning models. It is well-understood in the relevant art that there is typically a tradeoff between predictive accuracy and training speed, such that it is trivial to improve training speed by reducing predictive accuracy. Thus the real challenge is to improve training speed without sacrificing predictive accuracy through innovative model architectures. Accordingly, techniques that improve predictive accuracy without harming training speed, such as the techniques described herein, enable improving training speed given a constant predictive accuracy. In doing so, the techniques described herein improve efficiency and speed of training predictive machine learning models, thus reducing the number of computational operations needed and/or the amount of training data entries needed to train predictive machine learning models. Accordingly, the techniques described herein improve at least one of the computational efficiency, storage-wise efficiency, and speed of training machine learning models.
The process 400 begins at step/operation 402 when the predictive data analysis computing entity 106 identifies (e.g., receives) a plurality of input data objects. In some embodiments, the predictive data analysis computing entity 106 identifies a plurality of input data objects comprising a set of temporal sequences with respect to which one or more predictive data analysis operations are performed. A temporal sequence may comprise one or more classification features and a descriptive text feature associated with a plurality of related sequential events. In some embodiments, a temporal sequence describes an input data object comprising an instance xt={dt,et} at a given instance time t, for a given entity, where dt is representative of one or more classification features that describe the instance xt, and et is representative of a descriptive text feature associated with the instance xt. A temporal sequence may be descriptive of one or more events or actions associated with an entity for instance xt. For example, the instance may represent a patient admission on a given date and the given entity may represent a patient. As such, a temporal sequence may be representative of one or more events or actions for the instance xt. A temporal sequence may comprise images, text files, audio/video files, and application files that may be used to, for example, train a machine learning model. Classification features within a temporal sequence may be internally related and include correlations of various degrees that may be interpreted with various meanings.
In some embodiments, a set of temporal sequences describes plurality of temporal sequences En={x1n, x2n, . . . , x{T(n)}n} of a given entity n for T(n) number of instances xt for instance time 1-t. In some embodiments, the plurality of temporal sequences may be an ordered set of temporal sequences. As an example, the set of temporal sequences may comprise a patient's EHRs comprising a history of admissions where each of the plurality of temporal sequences may represent an admission within a time series. As such, a set of temporal sequences may be representative of a patient's condition over multiple admissions, procedures, and medications.
In some embodiments, a classification feature describes an attribute or characteristic associated with an input data object, such as a temporal sequence, which may be used for analysis and training of a machine learning model. In some embodiments, a classification feature may comprise descriptions, tags, or identifiers that classify or emphasize characteristics present in a body of data. As an example, classification features of an input data object may comprise clinical events including diagnoses, associated symptoms, or medical codes, e.g., International Statistical Classification of Diseases and Related Health Problems (ICD) codes, Current Procedural Terminology (CPT) codes, prescription (RX) codes.
In some embodiments, a descriptive text feature describes a textual description comprising a narrative description associated with a temporal sequence. For example, a descriptive text feature may comprise text describing one or more procedures performed on a patient.
An example of input data objects comprising a plurality of temporal sequences may be a collection of EHR data objects associated with a plurality of patients. The collection of EHR data objects may be extracted as a set of temporal sequences representative of a historical record associated with a given patient. Each temporal sequence may include one or more classification features and a descriptive text feature associated with a single admission of the given patient, such as clinical events including diagnoses, procedures, medications, etc. As such, an input data object may comprise a plurality of correlated relationships within its classification features for generating predictions.
An operational example of temporal sequences is depicted in
Referring back to
In some embodiments, co-occurrence probability describes a measurement of relationship between two variables. According to various embodiments of the present disclosure, the relationship may comprise a semantic proximity of the two variables. For example, co-occurrence frequency may comprise an above-chance frequency of classification features coinciding or existing within a body of data. As an example, a plurality of medical admission records may commonly include terms that describe certain comorbidities where there is a simultaneous presence of two or more medical conditions that are often related. As such, co-occurrence probability may reflect a statistical frequency in which certain terms are present within a single admission record.
At step/operation 406, the predictive data analysis computing entity 106 generates dynamic co-occurrence graphs for a given entity based on the global guidance correlation graph. Generating the dynamic co-occurrence graphs may include extracting temporal sequences from the input data object. According to some embodiments, dynamic co-occurrence graphs may be generated for each temporal sequence of a set of temporal sequences corresponding to the given entity. Dynamic co-occurrence graphs may comprise a plurality of semantic graphs including classification features associated with a plurality of temporal sequences over time. In some embodiments, a dynamic co-occurrence graph may be constructed based on global correlations from a global guidance correlation graph. In other words, a dynamic co-occurrence graph may comprise a projection of the global guidance correlation graph on a given temporal sequence.
According to various embodiments of the present disclosure, generating the dynamic co-occurrence graphs may comprise generating a sequence of adjacency matrices representative of a set of temporal sequences associated with a given entity, where each adjacency matrix comprises a fully connected graph including nodes that represent classification features corresponding to a given temporal sequence of the set of temporal sequences associated with the given entity. The adjacency matrices may further include edges and edge weights based on a global guidance correlation graph.
At step/operation 408, the predictive data analysis computing entity 106 generates initial embeddings based on the input data objects. In some embodiments, an attentional encoder decoder machine learning model processes classification features from the input data objects to generate the initial embeddings. As an example, generating the initial embeddings may comprise generating, for each temporal sequence within a set of temporal sequences for a given entity, an initial embedding corresponding to a given temporal sequence.
An initial embedding may comprise an initial representation of classification features from input data objects, where a classification feature matrix may be generated by processing the input data objects using an attentional encoder decoder machine learning model. In one embodiment, an initial embedding may comprise a translation of classification features from a given temporal sequence to a classification feature matrix comprising a tree hierarchy using an input embedding module of the attentional encoder decoder machine learning model.
For example, classification features of a given temporal sequence may be translated to a feature matrix Ht0∈R{|n
|nt| may represent a total number of classification features of an entity at the t-th instance time and d may represent feature dimension. We∈R{N×d} may represent an embedding to be learned, Ct0 may represent an initial embedding vector for the classification features at the t-th time, and N may represent the total number of classification features in a whole dataset.
In some embodiments, an attentional encoder decoder machine learning model describes parameters, hyperparameters, and/or defined operations of a machine learning model, where the machine learning model is configured to translate classification features into a classification feature matrix by processing input data objects. As described above, the attentional encoder decoder machine learning model may comprise at least an input embedding module that translates classification features from a temporal sequence to a classification feature matrix.
In some embodiments, the classification feature matrix may be created by using a tree-of-sequences LSTM network. In some embodiments, each classification feature may be associated with a description that describes the semantics of the classification features. For example, a classification feature, such as a diagnostic code, may be associated with a short text semantic description that describes the semantics of the diagnostic code. Other examples of such a description may include metadata or captioning. The descriptions may be extracted from the input data object or retrieved from a data source, e.g., a lookup database.
According to various embodiments of the present disclosure, the input embedding module may include a description encoder that generates a latent representation vector for a description corresponding to a classification feature. The description encoder may use a sequential LSTM network to encode descriptions of classification features. The input embedding module may further include a classification feature encoder that creates a tree-of-sequences LSTM network to capture hierarchal relationships among the classification features. Each node of the tree-of-sequences LSTM network may comprise an input vector based on the latent representation generated by the sequential LSTM.
In some embodiments, a LSTM network describes a recurrent neural network that stores a representation of classification features from sequences of input data objects, such as a set of temporal sequences, where long- and short-term information dependencies are preserved.
In some embodiments, a sequential LSTM network describes a LSTM network that learns a latent representation of classification features (which may reflect certain semantic information) and models a sequential structure among the classification features.
In some embodiments, a tree-of-sequences LSTM network describes a hierarchy of sequential LSTM networks. As described above, the tree-of-sequences LSTM network comprises classification features that are encoded by a sequential LSTM and applying a tree structure to the sequential LSTM to capture hierarchical relationships among the classification features. Each node of the tree-of-sequences LSTM network may comprise a vector including a latent representation based on the encoding by the sequential LSTM.
At step/operation 410, the predictive data analysis computing entity 106 generates historical node representations using the initial embeddings based on the dynamic co-occurrence graphs. In some embodiments, the initial embeddings are provided as input to a graph-attention augmented module. The initial embeddings, for example, may comprise initial representations of classification features for each of a plurality of temporal sequences within a set of temporal sequences for a given entity. The initial representations may be used by the graph-attention augmented module to build historical node representations based on adjacency information (e.g., information from node neighbors) from dynamic co-occurrence graphs.
According to various embodiments of the present disclosure, the graph-attention augmented module may generate historical node representations by using a plurality of dynamic co-occurrence graphs to update the initial embeddings over a progression through time corresponding to a set of temporal sequences. Each update to an initial embedding may comprise an encounter-level historical node representation corresponding to each temporal sequence within the set of temporal sequences. The updates may represent, for example, a progression of disease conditions for a patient. As such, a historical node representation may comprise an initial embedding that has been updated based on a dynamic co-occurrence graph. A historical node representation may be generated for each temporal sequence of a set of temporal sequences for a given entity. Accordingly, one or more of the historical node representations may be aggregated and used to perform predictive actions.
The graph-attention augmented module may comprise a tempo-spatial graph attention model (TSA) that generates the historical node representations using a graph-attention augmented temporal neural network. In some embodiments, a graph-attention augmented temporal neural network describes a two-layer graph-attention neural network for embedding historical node representations. In some embodiments, the graph-attention augmented temporal neural network model may update an initial embedding based on the dynamic co-occurrence graphs. According to one embodiment, for each time step representative of a given temporal sequence, a corresponding initial embedding may be provided to a TSA. The TSA may be trained with the initial embedding based on adjacency information (e.g., information from node neighbors) of a current temporal sequence of a current time step. The adjacency information may be provided to TSA as a dynamic co-occurrence graph corresponding to a given time step.
In some embodiments, a node representation may comprise a classification feature vector. The graph-attention augmented temporal neural network may construct encoded classification feature vectors that contain information of other co-occurrence classification features of a same temporal sequence with different degrees of correlation to obtain more comprehensive representation. As an example, at each embedding layer, the graph-attention augmented temporal neural network can embed a set of historical node representations Ht={h{t,1}, h{t,2}, . . . , h{t|n
where σ is a non-linear activation function, W and b are learnable parameters, and Ni is the set of neighboring nodes of node i in the graph.
In some embodiments, the graph-attention augmented temporal neural network may generate node representations as {H1, H2, . . . , Ht} at each time step which can capture the structural correlations, where Ht∈R{|c
Scores of classification features may be normalized to a probabilistic simplex using, for example, a Softmax operation according to the following equation:
Given normalized importance scores ã{t, i}, it can be used to apply weights to the representations of classification features comprising the single attentional vector x{t,i}=ã{t, i} h{t,i}.
A final historical node representation at instance time t may be represented by:
As such, the final historical node representation may be used to create embeddings for a machine learning classifier configured to generate one or more classification prediction outputs.
An operational example of building historical node representations is depicted in
According to the illustrated example depicted in
For a next time step T=2, an initial embedding for time step T=2 is provided to the TSA 606. TSA 606 trains on the initial embedding for time step T=2 based on adjacency information corresponding to time step T=2 and generates a historical node representation for time step T=2. This process is iteratively performed up to time step T=K, which may be representative of a present time temporal sequence (e.g., admission) where the TSA 606 may be used to perform a predictive function based on training of the initial embeddings with adjacency information over a plurality of time steps. The predictive function may comprise a classification task for predicting future conditions 608, such as class of disease prediction and link (e.g., edge with other ICD codes) prediction.
A machine learning framework 700 for generating historical node representations according to various embodiments of the present disclosure is depicted in
The initial embeddings may be provided to the graph-attention augmented modules 704A, 704B, . . . 704N. The initial embeddings may be used for model training by each of the respective graph-attention augmented modules 704A, 704B, . . . 704N to build respective historical node representations based on adjacency information (e.g., information from node neighbors) from respective dynamic co-occurrence graphs associated with each time T. In one embodiment, the graph-attention augmented modules 704A, 704B, . . . 704N may compute an attention coefficient for determining the importance of each of the neighbor's classification feature to a given historical node representation. The graph-attention augmented modules 704A, 704B, . . . 704N may be further configured to consider all classification features appearing in a set of temporal sequences, and embed all historical node representations that have been created up to time T=K to generate a final historical representation.
Node attention module 706 may append a single attention vector to historical node representations at time T=K. The single attention vector may comprise a combination of representations of all classification features associated with the historical node representations. An importance of each descriptive text feature of the temporal sequences to a given classification feature may also be determined. Importance scores may be generated based on the determined importance of each descriptive text feature to apply weighs to the classification features comprising the single attention vector.
Temporal dependency updating module 708 may be configured to model temporal evolution of each time step (e.g., temporal sequence) in the historical node representations at different time steps. For example, at each time step, when an initial embedding is updated, the corresponding historical node representation may selectively retain at least a portion of information from previous historical node representations. The temporal dependency updating module 708 may comprehensively capture the classification features appearing in a set of temporal sequences and generate classification feature vectors of the historical node representations that have been created to generate a final historical representation. An overall representation matrix may be incrementally generated by temporal dependency updating module 708 which may be transmitted to the classification module 710 to obtain one or more classification prediction outputs.
Returning to
According to one embedment, a classification prediction output may be generated based on classification features of the historical node representations where each node classification feature integrates the structural and temporal characteristics of the dynamic co-occurrence graphs. In one embodiment, the classification features from each historical node representation may be combined and used to generate final classification features. In another embodiment, the classification prediction output may comprise a multi-instance multi-label classification task. For example, the classification prediction output may be based on the plurality of prediction classification features for each time step corresponding to a plurality of historical node representations (e.g., temporal sequences associated with a set of temporal sequences).
According to some embodiments, generating the classification prediction output may comprise minimizing a loss function for optimizing the performance of the classification. For example, the loss function may be modeled as follows:
where ytj is the ground truth vector and ŷtj is the classification prediction output for a classification feature j at time instance t; Wj and bj may represent estimation parameters. Further ŷtj=1 if the j-th classification feature is predicted at t-th instance time step, otherwise it is zero. An algorithm, such as the Adam optimizer may be employed to minimize the loss function above.
At step/operation 414, the predictive data analysis computing entity 106 performs one or more prediction-based actions based on the classification prediction output. In some embodiments, performing the one or more prediction-based actions comprises performing one or more appointment scheduling operations and generating corresponding messages that are transmitted to client devices via an electronic communication system where the messages are rendered on one or more user interfaces. In another embodiment, performing the one or more prediction-based actions comprises generating one or more automated investigation operations and rendering a diagnosis on a user interface. In yet another embodiment, performing the one or more prediction-based actions comprises generating one or more automated audit operations based on the classification prediction output and rendering results of the one or more automated audit operations on a user interface.
The process 800 begins at step/operation 802. In some embodiments, the process 800 begins after one or more operations of another process, such as the step/operation 412 of the process 400 as depicted and described. Additionally or alternatively, in some embodiments, upon completion of the process 800, flow proceeds to one or more steps/operations of another process, such as the step/operation 414 of the process 400 as depicted and described. In other embodiments, the flow ends upon completion of the process 800.
At step/operation 802, the predictive data analysis computing entity 106 generates a plurality of prediction classification features. The plurality of prediction classification features may be generated based on a plurality of temporal sequences and each dynamic co-occurrence graph data object generated for a given entity based on a global guidance correlation graph, for example, as detailed in the above description relating to steps/operations 402 through 406. In some embodiments, a prediction classification feature describes an output of a machine learning model, such as a graph-attention augmented temporal neural network model, based on a given input data object. As an example, in some embodiments, prediction classification features may comprise medical codes, e.g., ICD codes, CPT codes, and RX codes that are generated for prediction on a set of temporal sequences. According to various embodiments of the present disclosure, a prediction classification feature may comprise a diagnostic code for a class of disease yt={0,1}L+1 at time instance t, where L is the number of classes of diseases considered, and all other diseases are represented as a single binary vector.
The graph-attention augmented temporal neural network model may comprise a plurality of embedding layers and is trained by, for each combination of a given temporal sequence t of T number of temporal sequences in the plurality of temporal sequences, a given non-initial embedding layer l of the plurality embedding layers, and a given classification feature i of the plurality of classification features: a) generating a historical node representation based on: (i) a prior-layer historical node representation for the given temporal sequence t and the given classification feature i as generated by a preceding embedding layer l−1, and (ii) neighbor nodes for a target node associated with the given classification feature i in the dynamic co-occurrence graph corresponding to the given temporal sequence t, and b) appending an attention vector comprising a node attention layer associated with the descriptive text feature to the historical node representation.
At step/operation 804, the predictive data analysis computing entity 106 generates one or more predicted edges in the global guidance correlation graph data object. In one embodiment, the one or more predicted edges may be connected to at least one node of the global guidance correlation graph data object associated with the plurality of prediction classification features. In some embodiments, a predicted edge describes a likely edge between two nodes of a global guidance correlation graph data object. The likely edge may be determined based on a vector generated by concatenating at least two nodes selected from the global guidance correlation graph data object. At least one of the two nodes may be associated with prediction classification features.
For example, a node u may be represented with its embedding gu (e.g., from the final historical node representation). N(u) may represent a set of nodes which form an edge with node u. To represent an edge (u, v) in a global correlation graph G, representations of the set of nodes in N(u) may be concatenated to generate a new vector z{u,v}=g{T,u}g{T,v} of size 2d+2d=4d. The combined representation may be fed into a fully-connected layer (FC) of a neural network, followed by a Softmax layer. The FC layer may output y{u,v} on an input of z{u,v}, and generate a prediction of an existence of a potential edge E{uv}.
The probability of an edge (u, v) being a reality may be obtained by the following equation:
A corresponding loss function may be minimized for existing edges E(G) on G by the following equation:
At step/operation 806, the predictive data analysis computing entity 106 determines one or more lowest common ancestor nodes of the plurality of prediction classification features from a classification feature graph based on the one or more predicted edges. In some embodiments, a lowest common ancestor node describes a root node of a smallest subtree within a classification feature graph including child nodes associated with at least one predicted edge within a global guidance correlation graph data object. The classification feature graph may comprise nodes and edges representative of a hierarchical relationship between a plurality of classification features.
At step/operation 808, the predictive data analysis computing entity 106 generates one or more link predictions based on the one or more lowest common ancestor nodes. In some embodiments, a link prediction describes a predicted classification of classification features identified from a classification feature graph based on one or more predicted edges. A link prediction may comprise an identification of a classification feature associated with a lowest common ancestor node. As such, a classification feature associated with a lowest common ancestor node may represent a class of classification features associated with its child nodes. For example, if a pair of prediction classification feature nodes in a classification feature graph belong to a same family, their common parent node may be identified as their classification. If a pair of prediction classification feature nodes are not connected (no subtree containing both), each of the prediction classification feature nodes may be treated as their own family.
Accordingly, as described above, various embodiments of the present disclosure make important technical contributions to improving predictive accuracy of predictive machine learning models, which improves training speed and training efficiency of machine learning models. This approach improves training speed and training efficiency of training predictive machine learning models. It is well-understood in the relevant art that there is typically a tradeoff between predictive accuracy and training speed, such that it is trivial to improve training speed by reducing predictive accuracy. Thus the real challenge is to improve training speed without sacrificing predictive accuracy through innovative model architectures. Accordingly, techniques that improve predictive accuracy without harming training speed, such as the techniques described herein, enable improving training speed given a constant predictive accuracy. In doing so, the techniques described herein improve efficiency and speed of training predictive machine learning models, thus reducing the number of computational operations needed and/or the amount of training data entries needed to train predictive machine learning models. Accordingly, the techniques described herein improve at least one of the computational efficiency, storage-wise efficiency, and speed of training machine learning models.
Furthermore, various embodiments of the present disclosure improve accuracy of machine learning models by pre-training models with classification features (e.g., clinical events) based on textual content as well as hierarchical structure among the classification features. As described herein, a collection of information, such as an electronic health record (EHR), may comprise a large number of classification features associated with a temporal structure. Existing methods for performing a prediction on a collection of information may include building a prediction model based on the classification features and making a prediction. However, existing methods are limited in their abilities in dealing with complex structural correlations and temporal dependencies of classification features in a temporal sequence (e.g., admission), which may impact classification predictions as well as temporal prediction of classification features.
However, in accordance with various embodiments of the present disclosure, a temporal-spatial approach may be used to capture temporal classification feature set (e.g., health) progression as well as relationships among different classification features over a period. This technique will lead to higher accuracy of performing predictions on input comprising a temporal sequence. In doing so, the techniques described herein improving efficiency and speed of training natural language processing machine learning models, thus reducing the number of computational operations needed and/or the amount of training data entries needed to train machine learning models. Accordingly, the techniques described herein improve at least one of the computational efficiency, storage-wise efficiency, and speed of training machine learning models.
Many modifications and other embodiments will come to mind to one skilled in the art to which this present disclosure pertains having the benefit of the teachings presented in the foregoing descriptions and the associated drawings. Therefore, it is to be understood that the present disclosure is not to be limited to the specific embodiments disclosed and that modifications and other embodiments are intended to be included within the scope of the appended claims. Although specific terms are employed herein, they are used in a generic and descriptive sense only and not for purposes of limitation.
