Patents Examined by Ying Yu Chen
  • Patent number: 10810512
    Abstract: Machine learning models used in medical diagnosis should be validated before being deployed in order to reduce the number of misdiagnoses. Validation processes presented here assess a performance of the machine learning model pre-deployment. In one or more examples, prior to the deployment of the machine learning model, the validation process assesses (1) whether a model achieves high enough performance to be deployed, and (2) that the process by which the performance metrics were computed was both sanitary and comprehensive. This pre-deployment validation helps prevent low-performing models from being deployed.
    Type: Grant
    Filed: March 21, 2018
    Date of Patent: October 20, 2020
    Assignee: Verily Life Sciences LLC
    Inventors: Peter Wubbels, Tyler Rhodes, Jin Zhang, Kira Whitehouse
  • Patent number: 10810497
    Abstract: A first element is extracted from a pair including a past inquiry and a past response, wherein the first element indicates that the past response shows an understanding of the past inquiry. A model is generated used to estimate a second element in a new inquiry based on the first element, wherein the second element indicates that a new response to the new inquiry shows an understanding of the new inquiry.
    Type: Grant
    Filed: March 9, 2016
    Date of Patent: October 20, 2020
    Assignee: International Business Machines Corporation
    Inventors: Emiko Takeuchi, Daisuke Takuma, Hirobumi Toyoshima
  • Patent number: 10810491
    Abstract: A visualization tool for machine learning models obtains metadata from a first training node at which a multi-layer machine learning model is being trained. The metadata includes a parameter of an internal layer of the model. The tool determines a plurality of metrics from the metadata, including respective loss function values corresponding to several training iterations of the model. The tool indicates the loss function values and the internal layer parameter values via a graphical interface.
    Type: Grant
    Filed: March 18, 2016
    Date of Patent: October 20, 2020
    Assignee: Amazon Technologies, Inc.
    Inventors: Wei Xia, Weixin Wu, Meng Wang, Ranju Das
  • Patent number: 10803349
    Abstract: Provided is a latent variable based taste determination method of determining including acquiring, by a computing apparatus, a response of the user on a plurality of questions associated with one of a plurality of areas including a gustatory area, a food ingredient area, and a restaurant environment area, or supporting another apparatus interacting with the computing apparatus to acquire the response; extracting, by the computing apparatus, independent latent variables representing taste load of the user in each of the plurality of areas, based on the response; extracting, by the computing apparatus, p hidden variables representing a correlation between the independent latent variables using a pretrained estimation model; and determining the taste of the user by comparing, by the computing apparatus, Euclidean distances between the hidden variables of the user and centroids of pre-grouped k taste groups. Here, each of p and k denotes a natural number of 1 or more.
    Type: Grant
    Filed: November 15, 2019
    Date of Patent: October 13, 2020
    Assignee: MABLIC CORPORATION
    Inventor: Yun Sik Choung
  • Patent number: 10789553
    Abstract: Examples of a digital orchestration system are provided. The system may obtain orchestration data on a real-time basis. The system may identify a plurality of events for offering a plurality of user services across a plurality of user interaction stages. The system may identify a plurality of actions associated with each of the plurality of events. The system may create a recipe associated with each of the plurality of actions. The system may identify and implement the associated recipe. The system may create an event sequence for each of the plurality of user interaction stages. The system may create a user service sequence comprising the plurality of user services associated with the event sequence. The system may generate a user experience result based on the event sequence and the user service sequence.
    Type: Grant
    Filed: December 30, 2019
    Date of Patent: September 29, 2020
    Assignee: ACCENTURE GLOBAL SOLUTIONS LIMITED
    Inventors: Suchit Vishnupant Bhatwadekar, Pankaj Shrikant Nikumb, Karthik Srinivasan
  • Patent number: 10789546
    Abstract: A computer-implemented method includes creating a classifier by: training a machine learning model using two or more tasks, wherein the tasks lie in two or more domains; including in the machine learning model at least one attribute common to at least two of said two or more domains; including in the machine learning model at least one latent feature that affects at least two of the two or more tasks that fall within one of the at least two domains; and constructing the classifier based on said machine learning model. The computer-implemented method further includes applying the classifier to at least one operational task.
    Type: Grant
    Filed: June 23, 2016
    Date of Patent: September 29, 2020
    Assignee: International Business Machines Corporation
    Inventors: Lakshminarayanan Krishnamurthy, Niyati Parameswaran
  • Patent number: 10789538
    Abstract: A computer-implemented method includes creating a classifier by: training a machine learning model using two or more tasks, wherein the tasks lie in two or more domains; including in the machine learning model at least one attribute common to at least two of said two or more domains; including in the machine learning model at least one latent feature that affects at least two of the two or more tasks that fall within one of the at least two domains; and constructing the classifier based on said machine learning model. The computer-implemented method further includes applying the classifier to at least one operational task.
    Type: Grant
    Filed: June 23, 2016
    Date of Patent: September 29, 2020
    Assignee: International Business Machines Corporation
    Inventors: Lakshminarayanan Krishnamurthy, Niyati Parameswaran
  • Patent number: 10783438
    Abstract: A method for on-device continual learning of a neural network which analyzes input data is provided to be used for smartphones, drones, vessels, or a military purpose. The method includes steps of: a learning device, (a) sampling new data to have a preset first volume, instructing an original data generator network, which has been learned, to repeat outputting synthetic previous data corresponding to a k-dimension random vector and previous data having been used for learning the original data generator network, such that the synthetic previous data has a second volume, and generating a batch for a current-learning; and (b) instructing the neural network to generate output information corresponding to the batch. The method can be performed by generative adversarial networks (GANs), online learning, and the like. Also, the present disclosure has effects of saving resources such as storage, preventing catastrophic forgetting, and securing privacy.
    Type: Grant
    Filed: June 17, 2019
    Date of Patent: September 22, 2020
    Assignee: StradVision, Inc.
    Inventors: Kye-Hyeon Kim, Yongjoong Kim, Insu Kim, Hak-Kyoung Kim, Woonhyun Nam, SukHoon Boo, Myungchul Sung, Donghun Yeo, Wooju Ryu, Taewoong Jang, Kyungjoong Jeong, Hongmo Je, Hojin Cho
  • Patent number: 10783452
    Abstract: A computer-implemented method is provided for learning a model corresponding to a target function that changes in time series. The method includes acquiring a time-series parameter that is a time series of input parameters including parameter values expressing the target function. The method further includes propagating propagation values, which are obtained by weighting parameters values at time points before one time point according to passage of the time points, to nodes in the model associated with the parameter values at the one time point. The method also includes calculating a node value of each node using each propagation value propagated to each node. The method additionally includes updating a weight parameter used for calculating the propagation values propagated to each node, using a difference between the target function at the one time point and a prediction function obtained by making a prediction from the node values of the nodes.
    Type: Grant
    Filed: April 28, 2017
    Date of Patent: September 22, 2020
    Assignee: INTERNATIONAL BUSINESS MACHINES CORPORATION
    Inventor: Hiroshi Kajino
  • Patent number: 10769161
    Abstract: Techniques are described for genomically defining digital genes encoding data visualization elements and potential incremental changes to the elements as the basis for a genetic selection process for automated generating of data visualizations. In one aspect, a method includes receiving set of input data. The method further includes generating digital genes that genomically define data visualization elements based on the input data, and that define potential incremental changes to the data visualization elements. The method further includes executing a genetic selection process with respect to one or more fitness functions on populations of candidate data visualizations that are based on the genomically defined data visualization elements. The method further includes outputting final data visualization output generated by the genetic selection process.
    Type: Grant
    Filed: November 3, 2015
    Date of Patent: September 8, 2020
    Assignee: International Business Machines Corporation
    Inventors: Daniel J. Rope, Graham J. Wills
  • Patent number: 10769162
    Abstract: Techniques are described for genomically defining digital genes encoding data visualization elements and potential incremental changes to the elements as the basis for a genetic selection process for automated generating of data visualizations. In one aspect, a method includes receiving set of input data. The method further includes generating digital genes that genomically define data visualization elements based on the input data, and that define potential incremental changes to the data visualization elements. The method further includes executing a genetic selection process with respect to one or more fitness functions on populations of candidate data visualizations that are based on the genomically defined data visualization elements. The method further includes outputting final data visualization output generated by the genetic selection process.
    Type: Grant
    Filed: March 16, 2017
    Date of Patent: September 8, 2020
    Assignee: International Business Machines Corporation
    Inventors: Daniel J. Rope, Graham J. Wills
  • Patent number: 10762424
    Abstract: Exemplary embodiments can maximize long-term value in a machine learning system. The system may employ an offline training process and an online training process. In the offline training process, an initial policy is learned to provide a warm start to the online training process. In the online training process, the system applies concurrent reinforcement learning across multiple environments, with the goal of learning efficient policies in real time from in-flight user data in one environment, and applying the learned policies to other environments. With the combination of offline training and online training, the system is able to improve initial performance through the warm start, while adapting to a changing context through concurrent reinforcement learning.
    Type: Grant
    Filed: September 11, 2018
    Date of Patent: September 1, 2020
    Assignee: SAS INSTITUTE INC.
    Inventors: Mohammad Reza Nazari, Afshin Oroojlooy Jadid, Mustafa Kabul
  • Patent number: 10762512
    Abstract: A system and method a method for providing for providing personalized transaction learning and tagging. The method may include tagging transactions associated with one or more financial accounts belonging to an account holder, whether the account holder be the primary, secondary, or a related account holder, such as a spouse, parent, guardian, and the like. The method may include linking all accounts belong to and/or associated with an account holder and receiving transaction data from each linked account, including, for example, transaction date, transaction time, transaction amount, merchant name, merchant location, merchant identifier, account number used in transaction, SKU-level transaction information, and/or other purchase identifiers (e.g., merchant-provided product/service name, account holder-provided product/service name, and the like). Once the system receives the transaction data, the system may query the account holder for input regarding the transaction data.
    Type: Grant
    Filed: November 2, 2015
    Date of Patent: September 1, 2020
    Assignee: CAPITAL ONE SERVICES, LLC
    Inventor: Moshe Benjamin
  • Patent number: 10755344
    Abstract: Embodiments of the invention are directed to a system, method, or computer program product for a framework processing for channel contacts data such as interaction data. In this way, the invention receives and processes raw interaction data from one or more deployed micro services. The framework subsequently provides for advanced operational execution not possible by the micro services. The processing of the interaction data includes machine learning predictive analytics framework and business rule application. As such, the system may learn from the data received at the micro servers in real-time and continue to monitor the raw interaction data based on the implemented business rules and machine learning to issue future alerts if necessary. Furthermore, the system may calculate analytics and perform advanced operation executions based on the demands/requests and/or rules. Finally the framework presents data in an interface picture format based on the calculated analytics.
    Type: Grant
    Filed: September 30, 2015
    Date of Patent: August 25, 2020
    Assignee: BANK OF AMERICA CORPORATION
    Inventors: Assim Syed Mohammad, Badri V. Mangalam, Prasanna Joshi, Sridhar M. Seetharaman
  • Patent number: 10755188
    Abstract: A system for predicting future behavior for a dynamic system comprises a processor configured to implement an artificial intelligence system implementing nonlinear modeling and forecasting processing for analyzing the dynamic system. The nonlinear modeling and forecasting processing configures the processor to generate a time series group of data from the dynamic system. The nonlinear modeling and forecasting processing further configures the processor to generate prediction values of future behavior of the dynamic system by using the nonlinear modeling and forecasting implemented on the artificial intelligence system on the time series group of data.
    Type: Grant
    Filed: October 16, 2018
    Date of Patent: August 25, 2020
    Assignee: NxGen Partners IP, LLC
    Inventor: Solyman Ashrafi
  • Patent number: 10748064
    Abstract: An artificial neural network and methods for performing computations on an artificial neural network include multiple neurons, including a layer of input neurons, one or more layers of hidden neurons, and a layer of output neurons. Arrays of weights are configured to accept voltage pulses from a first layer of neurons and to output current to a second layer of neurons during a feed forward operation. Each array of weights includes multiple resistive processing units having respective settable resistances.
    Type: Grant
    Filed: August 27, 2015
    Date of Patent: August 18, 2020
    Assignee: INTERNATIONAL BUSINESS MACHINES CORPORATION
    Inventors: Tayfun Gokmen, Seyoung Kim
  • Patent number: 10726330
    Abstract: System, method, and accelerator to process a convolutional neural network. In accordance therewith, a tile structure having input data values is loaded for a convolution layer. Each tile of the tile structure corresponds to a respective feature map in a set of input feature maps. The tile structure of each iteration represents a different subset of data values in the input feature maps. Intermediate data values associated with a subset of the data values of the input feature maps in the current intermediate tile structure are reused, when the intermediate data values of a previous tile structure overlap values to be computed in the current tile structure. Intermediate non-overlapping data values that are associated with the subset of the data values in the current tile structure are computed using associated filters having weight data values. Available reused intermediate data values and computed intermediate data values are buffered as intermediate data.
    Type: Grant
    Filed: October 11, 2017
    Date of Patent: July 28, 2020
    Assignee: The Research Foundation for The State University of New York
    Inventors: Michael Ferdman, Peter Milder, Manoj Alwani
  • Patent number: 10713565
    Abstract: Feature selection methods and processes that facilitate reduction of model components available for iterative modeling. It has been discovered that methods of eliminating model components that do not meaningfully contribute to a solution can be preliminarily discovered and discarded, thereby dramatically decreasing computational requirements in iterative programming techniques. This development unlocks the ability of iterative modeling to be used to solve complex problems that, in the past, would have required computation time on orders of magnitude too great to be useful.
    Type: Grant
    Filed: July 1, 2019
    Date of Patent: July 14, 2020
    Assignee: Liquid Biosciences, Inc.
    Inventor: Patrick Lilley
  • Patent number: 10713578
    Abstract: In one implementation, a method includes obtaining time series data. The time serious data includes a plurality of network utilization measurements. The plurality of network utilization measurements is indicative of a plurality of utilizations of one or more resources of a network resource at a plurality of times. The method also includes determining whether the time series data comprises a plurality of segments. Each segment of the plurality of segments is associated with a separate regression model and each segment includes a portion of the time series data. The method further includes identifying a current segment from the time series data when the time series data comprises the plurality of segments. The method further includes determining an estimated network utilization based on a current regression model associated with the current segment.
    Type: Grant
    Filed: April 29, 2015
    Date of Patent: July 14, 2020
    Assignee: CISCO TECHNOLOGY, INC.
    Inventors: Hao Hu, Preethi Natarajan, Nii Ako Ampa-Sowa, Stephen E. Jerman
  • Patent number: 10691997
    Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for augmenting neural networks to generate additional outputs. One of the systems includes a neural network and a sequence processing subsystem, wherein the sequence processing subsystem is configured to perform operations comprising, for each of the system inputs in a sequence of system inputs: receiving the system input; generating an initial neural network input from the system input; causing the neural network to process the initial neural network input to generate an initial neural network output for the system input; and determining, from a first portion of the initial neural network output for the system input, whether or not to cause the neural network to generate one or more additional neural network outputs for the system input.
    Type: Grant
    Filed: December 21, 2015
    Date of Patent: June 23, 2020
    Assignee: DeepMind Technologies Limited
    Inventors: Alexander Benjamin Graves, Ivo Danihelka, Gregory Duncan Wayne