Patent Application
20200250573
The present invention generally relates to artificial intelligence and machine learning, and more particularly to implementing a computer system task involving nonstationary streaming time-series data based on a bias-variance-based adaptive learning rate.
Machine learning models can be trained using one or more methods for optimizing objective functions (e.g., cost or loss functions). Such objective functions can be used to measure how well a machine learning model is performing by, e.g., assigning a cost to model errors associated with current parameter values. Objective function optimization methods can be used to train a wide range of models for implementing machine learning tasks, including but not limited to support vector machines, logistic regression and graphical models. In combination with a backpropagation algorithm, objective function optimization methods can be used to train neural networks for implementing machine learning tasks.
In accordance with an embodiment of the present invention, a system for implementing a computer system task involving nonstationary streaming time-series data is provided. The system includes a memory device for storing program code and at least one processor device operatively coupled to the memory device. The at least one processor device is configured to execute program code stored on the memory device to generate a parameter sequence including a plurality of parameters corresponding to respective iteration counts. The at least one processor device is configured to obtain a first parameter value corresponding to a given iteration count by calculating estimators of moments associated with an objective function corresponding to the given iteration count based on a second parameter value corresponding to a prior iteration count using a sequential mean tracking method, and obtaining the first parameter value by performing a step of a gradient descent method based on the calculated moments and the second parameter value. The at least one processor device is further configured to execute program code stored on the memory device to learn a time-series model based on the parameter sequence, and implement a computer system task using the time-series model.
In accordance with another embodiment of the present invention, a computer-implemented method for implementing a computer system task involving nonstationary streaming time-series data is provided. The method includes generating a parameter sequence including a plurality of parameters corresponding to respective iteration counts. Generating the parameter sequence includes obtaining a first parameter value corresponding to a given iteration count by calculating estimators of moments associated with an objective function corresponding to the given iteration count based on a second parameter value corresponding to a prior iteration count using a sequential mean tracking method, and obtaining the first parameter value by performing a step of a gradient descent method based on the calculated moments and the second parameter value. The method further includes learning a time-series model based on the parameter sequence, and implementing a computer system task using the time-series model.
These and other features and advantages will become apparent from the following detailed description of illustrative embodiments thereof, which is to be read in connection with the accompanying drawings.
The following description will provide details of preferred embodiments with reference to the following figures wherein:
The embodiments described herein can implement a computer system task involving nonstationary streaming time-series data by learning a time-series model in an online manner. A stationary time-series is a time-series with statistical properties (e.g., mean variance, autocorrelation) being constant over time, whereas a nonstationary time-series is defined as a time-series that is not stationary. For example, the embodiments described herein can apply a gradient descent method such as, e.g., stochastic gradient descent, that can handle nonstationary streaming time-series data to build and train a machine learning model. Gradient descent methods can find optimal parameters corresponding to optimal model performance by minimizing a corresponding objective function (e.g., cost or loss function) based on iteratively adjusting the parameters. The embodiments described herein can automatically tune a learning rate of the gradient descent method in a manner that reduces or eliminates biases in estimates of curvature and moments that can persist in memory, thereby increasing the speed that a processing device can automatically tune the learning rate for building and training a machine learning model. Accordingly, the embodiments described herein can improve the ability of a processing device to build and learn models for implementing computer system tasks (e.g., machine learning tasks) involving nonstationary streaming time-series.
Referring now to the drawings in which like numerals represent the same or similar elements and initially to
A first storage device 122 and a second storage device 124 are operatively coupled to system bus 102 by the I/O adapter 120. The storage devices 122 and 124 can be any of a disk storage device (e.g., a magnetic or optical disk storage device), a solid state magnetic device, and so forth. The storage devices 122 and 124 can be the same type of storage device or different types of storage devices.
A speaker 132 is operatively coupled to system bus 102 by the sound adapter 130. A transceiver 142 is operatively coupled to system bus 102 by network adapter 140. A display device 162 is operatively coupled to system bus 102 by display adapter 160.
A first user input device 152, a second user input device 154, and a third user input device 156 are operatively coupled to system bus 102 by user interface adapter 150. The user input devices 152, 154, and 156 can be any of a keyboard, a mouse, a keypad, an image capture device, a motion sensing device, a microphone, a device incorporating the functionality of at least two of the preceding devices, and so forth. Of course, other types of input devices can also be used, while maintaining the spirit of the present invention. The user input devices 152, 154, and 156 can be the same type of user input device or different types of user input devices. The user input devices 152, 154, and 156 are used to input and output information to and from system 100.
Parameter optimization (PO) component 170 may be operatively coupled to system bus 102. PO component 170 is configured to sample formulations within a formulation generation system as described in further detail below. PO component 170 can be implemented as a standalone special purpose hardware device, or may be implemented as software stored on a storage device. In the embodiment in which PO component 170 is software-implemented, although shown as a separate component of the computer system 100, PO component 170 can be stored on, e.g., the first storage device 122 and/or the second storage device 124. Alternatively, PO component 170 can be stored on a separate storage device (not shown).
Of course, the processing system 100 may also include other elements (not shown), as readily contemplated by one of skill in the art, as well as omit certain elements. For example, various other input devices and/or output devices can be included in processing system 100, depending upon the particular implementation of the same, as readily understood by one of ordinary skill in the art. For example, various types of wireless and/or wired input and/or output devices can be used. Moreover, additional processors, controllers, memories, and so forth, in various configurations can also be utilized as readily appreciated by one of ordinary skill in the art. These and other variations of the processing system 100 are readily contemplated by one of ordinary skill in the art given the teachings of the present invention provided herein.
It is to be understood that although this disclosure includes a detailed description on cloud computing, implementation of the teachings recited herein are not limited to a cloud computing environment. Rather, embodiments of the present invention are capable of being implemented in conjunction with any other type of computing environment now known or later developed.
Cloud computing is a model of service delivery for enabling convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, network bandwidth, servers, processing, memory, storage, applications, virtual machines, and services) that can be rapidly provisioned and released with minimal management effort or interaction with a provider of the service. This cloud model may include at least five characteristics, at least three service models, and at least four deployment models.
Characteristics are as follows:
On-demand self-service: a cloud consumer can unilaterally provision computing capabilities, such as server time and network storage, as needed automatically without requiring human interaction with the service's provider.
Broad network access: capabilities are available over a network and accessed through standard mechanisms that promote use by heterogeneous thin or thick client platforms (e.g., mobile phones, laptops, and PDAs).
Resource pooling: the provider's computing resources are pooled to serve multiple consumers using a multi-tenant model, with different physical and virtual resources dynamically assigned and reassigned according to demand. There is a sense of location independence in that the consumer generally has no control or knowledge over the exact location of the provided resources but may be able to specify location at a higher level of abstraction (e.g., country, state, or datacenter).
Rapid elasticity: capabilities can be rapidly and elastically provisioned, in some cases automatically, to quickly scale out and rapidly released to quickly scale in. To the consumer, the capabilities available for provisioning often appear to be unlimited and can be purchased in any quantity at any time.
Measured service: cloud systems automatically control and optimize resource use by leveraging a metering capability at some level of abstraction appropriate to the type of service (e.g., storage, processing, bandwidth, and active user accounts). Resource usage can be monitored, controlled, and reported, providing transparency for both the provider and consumer of the utilized service.
Service Models are as follows:
Software as a Service (SaaS): the capability provided to the consumer is to use the provider's applications running on a cloud infrastructure. The applications are accessible from various client devices through a thin client interface such as a web browser (e.g., web-based e-mail). The consumer does not manage or control the underlying cloud infrastructure including network, servers, operating systems, storage, or even individual application capabilities, with the possible exception of limited user-specific application configuration settings.
Platform as a Service (PaaS): the capability provided to the consumer is to deploy onto the cloud infrastructure consumer-created or acquired applications created using programming languages and tools supported by the provider. The consumer does not manage or control the underlying cloud infrastructure including networks, servers, operating systems, or storage, but has control over the deployed applications and possibly application hosting environment configurations.
Infrastructure as a Service (IaaS): the capability provided to the consumer is to provision processing, storage, networks, and other fundamental computing resources where the consumer is able to deploy and run arbitrary software, which can include operating systems and applications. The consumer does not manage or control the underlying cloud infrastructure but has control over operating systems, storage, deployed applications, and possibly limited control of select networking components (e.g., host firewalls).
Deployment Models are as follows:
Private cloud: the cloud infrastructure is operated solely for an organization. It may be managed by the organization or a third party and may exist on-premises or off-premises.
Community cloud: the cloud infrastructure is shared by several organizations and supports a specific community that has shared concerns (e.g., mission, security requirements, policy, and compliance considerations). It may be managed by the organizations or a third party and may exist on-premises or off-premises.
Public cloud: the cloud infrastructure is made available to the general public or a large industry group and is owned by an organization selling cloud services.
Hybrid cloud: the cloud infrastructure is a composition of two or more clouds (private, community, or public) that remain unique entities but are bound together by standardized or proprietary technology that enables data and application portability (e.g., cloud bursting for load-balancing between clouds).
A cloud computing environment is service oriented with a focus on statelessness, low coupling, modularity, and semantic interoperability. At the heart of cloud computing is an infrastructure that includes a network of interconnected nodes.
Referring now to
Referring now to
Hardware and software layer 360 includes hardware and software components. Examples of hardware components include: mainframes 361; RISC (Reduced Instruction Set Computer) architecture based servers 362; servers 363; blade servers 364; storage devices 365; and networks and networking components 366. In some embodiments, software components include network application server software 367 and database software 368.
Virtualization layer 370 provides an abstraction layer from which the following examples of virtual entities may be provided: virtual servers 371; virtual storage 372; virtual networks 373, including virtual private networks; virtual applications and operating systems 374; and virtual clients 375.
In one example, management layer 380 may provide the functions described below. Resource provisioning 381 provides dynamic procurement of computing resources and other resources that are utilized to perform tasks within the cloud computing environment. Metering and Pricing 382 provide cost tracking as resources are utilized within the cloud computing environment, and billing or invoicing for consumption of these resources. In one example, these resources may include application software licenses. Security provides identity verification for cloud consumers and tasks, as well as protection for data and other resources. User portal 383 provides access to the cloud computing environment for consumers and system administrators. Service level management 384 provides cloud computing resource allocation and management such that required service levels are met. Service Level Agreement (SLA) planning and fulfillment 385 provide pre-arrangement for, and procurement of, cloud computing resources for which a future requirement is anticipated in accordance with an SLA.
Workloads layer 390 provides examples of functionality for which the cloud computing environment may be utilized. Examples of workloads and functions which may be provided from this layer include: mapping and navigation 391; software development and lifecycle management 392; virtual classroom education delivery 393; data analytics processing 394; transaction processing 395; and telematics data processing 396.
The present invention may be a system, a method, and/or a computer program product at any possible technical detail level of integration. The computer program product may include a computer readable storage medium (or media) having computer readable program instructions thereon for causing a processor to carry out aspects of the present invention.
The computer readable storage medium can be a tangible device that can retain and store instructions for use by an instruction execution device. The computer readable storage medium may be, for example, but is not limited to, an electronic storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. A non-exhaustive list of more specific examples of the computer readable storage medium includes the following: a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or Flash memory), a static random access memory (SRAM), a portable compact disc read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanically encoded device such as punch-cards or raised structures in a groove having instructions recorded thereon, and any suitable combination of the foregoing. A computer readable storage medium, as used herein, is not to be construed as being transitory signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through a waveguide or other transmission media (e.g., light pulses passing through a fiber-optic cable), or electrical signals transmitted through a wire.
Computer readable program instructions described herein can be downloaded to respective computing/processing devices from a computer readable storage medium or to an external computer or external storage device via a network, for example, the Internet, a local area network, a wide area network and/or a wireless network. The network may comprise copper transmission cables, optical transmission fibers, wireless transmission, routers, firewalls, switches, gateway computers and/or edge servers. A network adapter card or network interface in each computing/processing device receives computer readable program instructions from the network and forwards the computer readable program instructions for storage in a computer readable storage medium within the respective computing/processing device.
Computer readable program instructions for carrying out operations of the present invention may be assembler instructions, instruction-set-architecture (ISA) instructions, machine instructions, machine dependent instructions, microcode, firmware instructions, state-setting data, or either source code or object code written in any combination of one or more programming languages, including an object oriented programming language such as SMALLTALK, C++ or the like, and conventional procedural programming languages, such as the “C” programming language or similar programming languages. The computer readable program instructions may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server. In the latter scenario, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or the connection may be made to an external computer (for example, through the Internet using an Internet Service Provider). In some embodiments, electronic circuitry including, for example, programmable logic circuitry, field-programmable gate arrays (FPGA), or programmable logic arrays (PLA) may execute the computer readable program instructions by utilizing state information of the computer readable program instructions to personalize the electronic circuitry, in order to perform aspects of the present invention.
Aspects of the present invention are described herein with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer readable program instructions.
These computer readable program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions/acts specified in the flowchart and/or block diagram block or blocks. These computer readable program instructions may also be stored in a computer readable storage medium that can direct a computer, a programmable data processing apparatus, and/or other devices to function in a particular manner, such that the computer readable storage medium having instructions stored therein comprises an article of manufacture including instructions which implement aspects of the function/act specified in the flowchart and/or block diagram block or blocks.
The computer readable program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable apparatus or other device to produce a computer implemented process, such that the instructions which execute on the computer, other programmable apparatus, or other device implement the functions/acts specified in the flowchart and/or block diagram block or blocks.
The flowchart and block diagrams in the Figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of instructions, which comprises one or more executable instructions for implementing the specified logical function(s). In some alternative implementations, the functions noted in the blocks may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems that perform the specified functions or acts or carry out combinations of special purpose hardware and computer instructions.
Reference in the specification to “one embodiment” or “an embodiment” of the present invention, as well as other variations thereof, means that a particular feature, structure, characteristic, and so forth described in connection with the embodiment is included in at least one embodiment of the present invention. Thus, the appearances of the phrase “in one embodiment” or “in an embodiment”, as well any other variations, appearing in various places throughout the specification are not necessarily all referring to the same embodiment.
It is to be appreciated that the use of any of the following “/”, “and/or”, and “at least one of”, for example, in the cases of “A/B”, “A and/or B” and “at least one of A and B”, is intended to encompass the selection of the first listed option (A) only, or the selection of the second listed option (B) only, or the selection of both options (A and B). As a further example, in the cases of “A, B, and/or C” and “at least one of A, B, and C”, such phrasing is intended to encompass the selection of the first listed option (A) only, or the selection of the second listed option (B) only, or the selection of the third listed option (C) only, or the selection of the first and the second listed options (A and B) only, or the selection of the first and third listed options (A and C) only, or the selection of the second and third listed options (B and C) only, or the selection of all three options (A and B and C). This may be extended, as readily apparent by one of ordinary skill in this and related arts, for as many items listed.
With reference to
Although the system architecture 400 is shown being implemented within the context of a factory 410, the system architecture 400 can be implemented within any suitable context. Examples of other contexts, but are not limited to, consumer applications (e.g., “smart home” systems including smart home objects providing home automation), and commercial applications (e.g., “smart healthcare” systems including smart medical objects providing remote health monitoring).
As shown, the system architecture 400 further includes an anomaly detection system 430 for performing an anomaly detection task within the factory 410. In performing this task, a statistical model can be used to learn patterns of sensor inputs from the devices 420-1 through 420-N. Such a model can be applied to check whether upcoming sensor inputs are predictable or not and, if the upcoming sensor inputs are not predictable, an anomaly is detected.
For example, the anomaly detection system 430 can include a model learning component 432 for learning a machine learning model for detecting anomalies, and an anomaly detector 434 for performing an anomaly detection task by applying the learned machine learning model to sensor inputs. The anomaly detector 434 can be further configured to automatically generate an alert when an anomaly is detected based on the machine learning model.
If the sensor inputs are non-stationary, it can be difficult to implement the anomaly detection system 430 due at least in part to the inability to track changes and reliance on tuning using expert knowledge. As will be further described in detail below with reference to
With reference to
Illustratively, the computer system task can include learning a time-series model, pθ(Xt+1|xt), where θ is a parameter, xt=x1, x2, . . . , xt is a set of past observations, and once it is observed that Xt+1=xt+1, the parameter θ is updated. More specifically, the system/method 500 can be used to learn a series of model parameters θt*, such that:
so as to adapt to the changing environment, since the optimal parameter can change in the nonstationary environment. In equation (1), (θ; Xt) refers to an objective function,
refers to the expectation, and
(θ) refers to the true objective function.
Generally, an objective function is a function that is optimized during training during a machine learning task. In one embodiment, the series of objective functions includes a series of loss or cost functions.
As will be described in further detail below, the system/method 500 can solve the optimization above (e.g., by approximation), because the past observations are represented as a series of realizations xt=x1, x2, . . . , xt and the true objective function (θ) is not available (the true objective function cannot be calculated without the true distribution pθ(Xt+1|xt), which we wish to learn). The objective function can be minimized by θt=θt−1−ηt*∇
(θt−1) using, e.g., stochastic gradient descent with an optimal learning rate
where gt=∇t(θt−1), vt=(∇
t(θt−1))2, and ht=∇2
t(θt−1), and
SMT refers to an online process that, at each iteration count or time step t, can receive an observation zt and output an estimation of the mean at iteration count t, {circumflex over (z)}t. More specifically, SMT can update the estimator {circumflex over (z)}t using an update rule, such as, e.g., {circumflex over (z)}t→ât{circumflex over (z)}t−1+(1−ât)zt, where the weight is defined as
and is computed using the following estimators: {circumflex over (σ)}=t−{circumflex over (z)}t2, {circumflex over (γ)}t2=
t−|{circumflex over (z)}t|2, {circumflex over (δ)}=|{circumflex over (z)}t−{circumflex over (z)}t|, and {circumflex over (z)}t=ât−1{circumflex over (z)}t−1+(1−ât−1){circumflex over (z)}t,
t=ât−1
t−1+(1−ât−1)|zt|2,
t=ât−1
t−1+(1−ât−1)|{circumflex over (z)}5|2. The estimators can be initialized as, e.g., â032 0.5, {circumflex over (z)}1=z1, {circumflex over (z)}1=z1
1=|z1|2,
1=|z1|2.
In accordance with the embodiments described herein, SMT can be used to estimate
At block 502, a set of inputs is received. The set of inputs can include a series of objective functions lt and an initial parameter value θ0.
At block 504, the iteration count t, the estimators
At block 506, the iteration count t is updated (e.g., t→t+1).
At block 508, a gradient gt and curvature ht of the objective function ; are computed. The gradient gt and curvature ht can be computed based on the parameter value. For example, gt→∇
t(θt−1); ht→∇2
t(θt−1).
At block 510, SMT is used to calculate the estimators of moments based on the gradient and the curvature. More specifically, for each random variable z in (gt,vt,ht) and its estimator {circumflex over (z)}, the following can be performed:
1. Estimate trVar(z), trVar({circumflex over (z)}), |E{circumflex over (z)}−Ez|, where Var( )refers to variance, E( ) refers to expectation, and tr refers to the trace operator. For example, letting {circumflex over (z)}t=ât−1{circumflex over (z)}t−1+(1−ât−1){circumflex over (z)}t, t=ât−1
t−1+(1−ât−1)|zt|2,
t=ât−1
t−1+(1−ât−1)|{circumflex over (z)}t|2, the statistics of our interest, trVar(z), trVar({circumflex over (z)}), E(E{circumflex over (z)}−Ez), can be estimated using SMT as {circumflex over (σ)}t2=
t−|{circumflex over (z)}t|2, γt2=
t−|{circumflex over (z)}5|2 , δt=|{circumflex over (z)}t−{circumflex over (z)}t|, respectively.
where the statistics are replaced with the estimators.
At block 512, a learning rate ηt is estimated based on the calculated moments. The learning rate determines how “fast” or “slow” the gradient descent method moves toward the solution. For the gradient descent method to reach the minimum of the objective function, the learning rate needs to be set to an appropriate value. For example, if the learning rate is set too high, the steps the gradient descent can bounce around and not reach the minimum, and if the learning rate is set too low, it may take too much time and resources to reach the minimum.
For example, the learning rate can be defined as
where
where
The sequential mean tracking performed at block 510 produces approximate solutions for the optimal bias-variance trade-off problem of the calculation of moments in an efficient manner. By using sequential mean tracking to calculate the moments, the approximation error of an optimal learning rate is reduced. Accordingly, the sequential mean tracking performed at block 510 can improve the ability of a processing device to estimate the learning rate at block 512, thereby improving the ability of a processing device to implement a machine learning task involving optimizing a model based on the learning rate (e.g., model and/or neural network training).
At block 514, the parameter value is updated by performing a step of a gradient descent method. For example, the parameter value θt−1 can be updated to θt. Generally, gradient descent is used to find parameter values that minimize the objective function by iteratively adjusting the parameter values. For example, the gradient descent method can include stochastic gradient descent, which is scalable in handling streaming data (e.g., nonstationary streaming time-series data). Using stochastic gradient descent, the parameter value θt−1 can be updated to θt as θt→θt−1−ηtgt.
At block 516, it is determined if t=T. If not, the process reverts back to block 506 to update the iteration count. Otherwise, a parameter value sequence {θt}t=1T is output. In an alternative embodiment, θt is output for each iteration count t, regardless of whether t=T.
At block 518, a time-series model is learned in an online manner based on the parameter sequence. The time-series model can be learned in an online manner.
At block 520, a computer system task is implemented using the time-series model. In one embodiment, the computer system task can include a machine learning task. For example, as described above with reference to
Having described preferred embodiments of systems and methods of implementing a computer system task involving nonstationary streaming time-series data based on a bias-variance-based adaptive learning rate (which are intended to be illustrative and not limiting), it is noted that modifications and variations can be made by persons skilled in the art in light of the above teachings. It is therefore to be understood that changes may be made in the particular embodiments disclosed which are within the scope of the invention as outlined by the appended claims. Having thus described aspects of the invention, with the details and particularity required by the patent laws, what is claimed and desired protected by Letters Patent is set forth in the appended claim.
