Patents Examined by Luis Sitiriche
  • Patent number: 9990591
    Abstract: Invoking an agent during a dialog between a user and an automated assistant. Some implementations are directed to receiving, during a human-to-automated assistant dialog, natural language input of the user that indicates a desire to engage an agent, but that fails to indicate a particular agent to be engaged. Those implementations are further directed to selecting a particular agent from a plurality of available agents, and transmitting an invocation request to the selected particular agent. In some implementations an agent selection model can be utilized in selecting the particular agent, such as a machine learning model. The machine learning model can be trained to enable generation of output that indicates, for each of a plurality of available agents (and optionally intent(s) for those agents), a probability that the available agent (and optionally intent) will generate appropriate responsive content.
    Type: Grant
    Filed: April 18, 2017
    Date of Patent: June 5, 2018
    Assignee: GOOGLE LLC
    Inventors: Ilya Gennadyevich Gelfenbeyn, Artem Goncharuk, Pavel Sirotin
  • Patent number: 9974226
    Abstract: A method begins by agriculture equipment collecting current on-site gathered agriculture data regarding an agriculture region and sending at least a representation of the current on-site gathered agriculture data to a host device. The method continues with the host device processing one or more of the at least a representation of the current on-site gathered agriculture data, current off-site gathered agriculture data, historical on-site gathered agriculture data, historical off-site gathered agriculture data, and historical analysis of agriculture predictions regarding the agriculture region to produce a current agriculture prediction for the agriculture region. The method continues with the host device generating an agriculture prescription regarding at least a portion of the agriculture region based on the current agriculture prediction and sending the agriculture prescription to one or more of the agriculture equipment.
    Type: Grant
    Filed: April 20, 2015
    Date of Patent: May 22, 2018
    Assignee: The Climate Corporation
    Inventors: Craig Eugene Rupp, A. Corbett S. Kull, Steve Richard Pitstick, Patrick Lee Dumstorff
  • Patent number: 9971974
    Abstract: Provided are methods and systems for knowledge discovery utilizing knowledge profiles.
    Type: Grant
    Filed: August 25, 2016
    Date of Patent: May 15, 2018
    Assignee: Elsevier, Inc.
    Inventors: Edwin Adriaansen, Bob J. A. Schijvenaars
  • Patent number: 9959940
    Abstract: A computer implemented system and method provides a volume of activation (VOA) estimation model that receives as input two or more electric field values of a same or different data type at respective two or more positions of a neural element and determines based on such input an activation status of the neural element. A computer implemented system and method provides a machine learning system that automatically generates a computationally inexpensive VOA estimation model based on output of a computationally expensive system.
    Type: Grant
    Filed: March 10, 2017
    Date of Patent: May 1, 2018
    Assignee: Boston Scientific Neuromodulation Corporation
    Inventors: Michael A. Moffitt, G. Karl Steinke
  • Patent number: 9959504
    Abstract: Certain relationships representing material insights are identified from among a set of discovered relationships. Cognitive discovery of relationships in a knowledge base, or corpus, are ranked according to one or more metrics indicative of material insights, including recentness and degree of alignment.
    Type: Grant
    Filed: December 2, 2015
    Date of Patent: May 1, 2018
    Assignee: International Business Machines Corporation
    Inventors: John B. Gordon, John P. Hogan, Sanjay F. Kottaram
  • Patent number: 9953271
    Abstract: Technologies are generally described for systems, devices and methods relating to determining weights in a machine learning environment. In some examples, a training distribution of training data may be identified, information about a test distribution of test data, and a coordinate of the training data and the test data may be identified. Differences between the test distribution and the training distribution may be determined, for the coordinate. A weight importance parameter may be identified, for the coordinate. A processor may calculate weights based on the differences, and based on the weight importance parameter. The weights may be adapted to cause the training distribution to conform to the test distribution at a degree of conformance. The degree of conformance may be based on the weight importance parameter.
    Type: Grant
    Filed: August 5, 2014
    Date of Patent: April 24, 2018
    Assignee: CALIFORNIA INSTITUTE OF TECHNOLOGY
    Inventors: Yaser Said Abu-Mostafa, Carlos Roberto Gonzalez
  • Patent number: 9946970
    Abstract: Embodiments described herein are directed to methods and systems for performing neural network computations on encrypted data. Encrypted data is received from a user. The encrypted data is encrypted with an encryption scheme that allows for computations on the ciphertext to generate encrypted results data. Neural network computations are performed on the encrypted data, using approximations of neural network functions to generate encrypted neural network results data from encrypted data. The approximations of neural network functions can approximate activation functions, where the activation functions are approximated using polynomial expressions. The encrypted neural network results data are communicated to the user associated with the encrypted data such that the user decrypts the encrypted data based on the encryption scheme. The functionality of the neural network system can be provided using a cloud computing platform that supports restricted access to particular neural networks.
    Type: Grant
    Filed: November 7, 2014
    Date of Patent: April 17, 2018
    Assignee: Microsoft Technology Licensing, LLC
    Inventors: Ran Gilad-Bachrach, Thomas William Finley, Mikhail Bilenko, Pengtao Xie
  • Patent number: 9934338
    Abstract: Building models and predicting operational outcomes of a drilling operation. At least some of the illustrative embodiments are methods including: gathering sensor data regarding offset wells and context data regarding the offset wells, and placing the sensor data and context data into a data store; creating a reduced data set by identifying a correlation between data in the data store and an operational outcome in a drilling operation; creating a model based on the reduced data set; and predicting the operational outcome based on the model.
    Type: Grant
    Filed: June 7, 2013
    Date of Patent: April 3, 2018
    Assignee: LANDMARK GRAPHICS CORPORATION
    Inventors: Olivier Germain, Keshava P. Rangarajan, Amit K. Singh, Hermanus Teunissen, Ram N. Adari
  • Patent number: 9928468
    Abstract: Various embodiments train a prediction model for predicting a label to be allocated to a prediction target explanatory variable set. In one embodiment, one or more sets of training data are acquired. Each of the one or more sets of training data includes at least one set of explanatory variables and a label allocated to the at least one explanatory variable set. A plurality of explanatory variable subsets is extracted from the at least one set of explanatory variables. A prediction model is trained utilizing the training data. The plurality of explanatory variable subsets is reflected on a label predicted by the prediction model to be allocated to a prediction target explanatory variable set with each of the plurality of explanatory variable subsets weighted respectively.
    Type: Grant
    Filed: January 31, 2017
    Date of Patent: March 27, 2018
    Assignee: International Business Machines Corporation
    Inventors: Takayuki Katsuki, Yuma Shinohara
  • Patent number: 9922292
    Abstract: Various embodiments train a prediction model for predicting a label to be allocated to a prediction target explanatory variable set. In one embodiment, one or more sets of training data are acquired. Each of the one or more sets of training data includes at least one set of explanatory variables and a label allocated to the at least one explanatory variable set. A plurality of explanatory variable subsets is extracted from the at least one set of explanatory variables. A prediction model is trained utilizing the training data. The plurality of explanatory variable subsets is reflected on a label predicted by the prediction model to be allocated to a prediction target explanatory variable set with each of the plurality of explanatory variable subsets weighted respectively.
    Type: Grant
    Filed: September 22, 2015
    Date of Patent: March 20, 2018
    Assignee: International Business Machines Corporation
    Inventors: Takayuki Katsuki, Yuma Shinohara
  • Patent number: 9916533
    Abstract: A method for ingesting a plurality of content according to a statistical similarity of at least one portion of the ingested plurality of content into an information handling system capable of answering questions, whereby the ingested plurality of content is based on a received topic and ingesting the plurality of content comprises ingesting a plurality of documents associated with the received topic is provided. The method may include determining at least one similarity between each document based on a similarity criteria. The method may also include applying a statistical model to characterize the determined at least one similarity between each document. The method may further include creating at least one pair-wise link for each document. The method may additionally include mapping the created at least one pair-wise link. The method may include generating a plurality of rules for ingesting a plurality of additional content.
    Type: Grant
    Filed: March 10, 2015
    Date of Patent: March 13, 2018
    Assignee: International Business Machines Corporation
    Inventors: Paul R. Bastide, Matthew E. Broomhall, Robert E. Loredo, Dale M. Schultz
  • Patent number: 9916534
    Abstract: A method for ingesting a plurality of content according to a statistical similarity of at least one portion of the ingested plurality of content into an information handling system capable of answering questions, whereby the ingested plurality of content is based on a received topic and ingesting the plurality of content comprises ingesting a plurality of documents associated with the received topic is provided. The method may include determining at least one similarity between each document based on a similarity criteria. The method may also include applying a statistical model to characterize the determined at least one similarity between each document. The method may further include creating at least one pair-wise link for each document. The method may additionally include mapping the created at least one pair-wise link. The method may include generating a plurality of rules for ingesting a plurality of additional content.
    Type: Grant
    Filed: September 9, 2015
    Date of Patent: March 13, 2018
    Assignee: International Business Machines Corporation
    Inventors: Paul R. Bastide, Matthew E. Broomhall, Robert E. Loredo, Dale M. Schultz
  • Patent number: 9858534
    Abstract: Technologies are generally described for systems, devices and methods relating to a machine learning environment. In some examples, a processor may identify a training distribution of a training data. The processor may identify information about a test distribution of a test data. The processor may identify a coordinate of the training data and the test data. The processor may determine, for the coordinate, differences between the test distribution and the training distribution. The processor may determine weights based on the differences. The weights may be adapted to cause the training distribution to conform to the test distribution when the weights are applied to the training distribution.
    Type: Grant
    Filed: August 5, 2014
    Date of Patent: January 2, 2018
    Assignee: CALIFORNIA INSTITUTE OF TECHNOLOGY
    Inventors: Yaser Said Abu-Mostafa, Carlos Roberto Gonzalez
  • Patent number: 9792412
    Abstract: A computer implemented system and method provides a volume of activation (VOA) estimation model that receives as input two or more electric field values of a same or different data type at respective two or more positions of a neural element and determines based on such input an activation status of the neural element. A computer implemented system and method provides a machine learning system that automatically generates a computationally inexpensive VOA estimation model based on output of a computationally expensive system.
    Type: Grant
    Filed: August 22, 2013
    Date of Patent: October 17, 2017
    Assignee: Boston Scientific Neuromodulation Corporation
    Inventors: Michael A. Moffitt, G. Karl Steinke
  • Patent number: 9785847
    Abstract: Apparatus, systems, and methods for analyzing data are described. The data can be analyzed using a hierarchical structure. One such hierarchical structure can comprise a plurality of layers, where each layer performs an analysis on input data and provides an output based on the analysis. The output from lower layers in the hierarchical structure can be provided as inputs to higher layers. In this manner, lower layers can perform a lower level of analysis (e.g., more basic/fundamental analysis), while a higher layer can perform a higher level of analysis (e.g., more complex analysis) using the outputs from one or more lower layers. In an example, the hierarchical structure performs pattern recognition.
    Type: Grant
    Filed: November 22, 2013
    Date of Patent: October 10, 2017
    Assignee: Micron Technology, Inc.
    Inventor: Paul Dlugosch
  • Patent number: 9767418
    Abstract: Events may be identified by storing information in response to activating an event stamp function. As a result of activating the event function, the information collected may immediately be compared to event information in a database. Alternatively, the information collected may later be compared to event information in a database. One or more candidates for the event of interest may be automatically or manually retrieved, and the user may decide whether a candidate event of the one or more candidates correspond to the event of interest. Alternatively, a purchase of an item related to the event may be automatically made in response to activating the event stamp.
    Type: Grant
    Filed: October 29, 2012
    Date of Patent: September 19, 2017
    Assignee: Proximity Grid, Inc.
    Inventor: John H. Reimer
  • Patent number: 9710755
    Abstract: A system and method for predicting search term popularity is disclosed herein. A database system may comprise a first database cluster H and a second database cluster L. A machine learning algorithm is trained to create a predictive model. Thereafter, for each record in a database system, the predictive model is used to calculate a probability of the record being accessed. If the calculated probability of the record being accessed is greater than a threshold value, then the record in the first database cluster H; otherwise, the record is placed in the second database cluster L. Training the machine learning algorithm comprises inputting a training feature vector associated with the record into the machine learning algorithm, inputting a cost vector into the machine learning algorithm, and iteratively operating the machine learning algorithm on each record in the set of records to create a predictive model. Other embodiments are also disclosed herein.
    Type: Grant
    Filed: September 26, 2014
    Date of Patent: July 18, 2017
    Assignee: WAL-MART STORES, INC.
    Inventors: Varun Srivastava, Yiye Ruan, Yan Zheng
  • Patent number: 9704105
    Abstract: Various embodiments select features from a feature space. In one embodiment, a set of training samples and a set of test samples are received. A first centered Gram matrix of a given dimension is determined for each of a set of feature vectors that include at least one of the set of training samples and at least one of the set of test samples. A second centered Gram matrix of the given dimension is determined for a target value vector that includes target values from the set of training samples. A set of columns and rows associated with the at least one of the test samples in the second centered Gram matrix is set to 0. A subset of features is selected from a set of features based on the first and second centered Gram matrices.
    Type: Grant
    Filed: May 16, 2016
    Date of Patent: July 11, 2017
    Assignee: International Business Machines Corporation
    Inventors: Dan He, Laxmi P. Parida, Irina Rish
  • Patent number: 9704102
    Abstract: A mechanism for discerning user preferences for categories of provided content. A computer receives response data including a set of preference values that have been assigned to content items by content users. Output data is computed based on the response data using a latent factor model. The output data includes at least: an association matrix that defines K concepts associated with the content items, wherein K is smaller than the number of the content items, wherein, for each of the K concepts, the association matrix defines the concept by specifying strengths of association between the concept and the content items; and a concept-preference matrix including, for each content user and each of the K concepts, an extent to which the content user prefers the concept. The computer may display a visual representation of the association strengths in the association matrix and/or the extents in the concept-preference matrix.
    Type: Grant
    Filed: March 15, 2014
    Date of Patent: July 11, 2017
    Assignee: William Marsh Rice University
    Inventors: Richard G. Baraniuk, Andrew S. Lan, Christoph E. Studer, Andrew E. Waters
  • Patent number: 9697470
    Abstract: A method is provided for determining one or more causes for variability in data. The method includes selecting a first range of a multivariate model output data on a user interface and employing a computing system, operatively coupled to the user interface, to determine one or more process data causing a variability of the multivariate model output data in the first range when compared to a second range of the multivariate model output data. At least some of the process data includes data derived from a physical measurement of a process variable.
    Type: Grant
    Filed: April 16, 2014
    Date of Patent: July 4, 2017
    Assignee: Applied Materials, Inc.
    Inventors: Jimmy Iskandar, Bradley D. Schulze, Kommisetti Subrahmanyam, Haw-Jyue Luo