Patents by Inventor Keith Bonawitz

Keith Bonawitz has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).

  • Patent number: 11488054
    Abstract: The present disclosure provides systems and methods for distributed training of machine learning models. In one example, a computer-implemented method is provided for training machine-learned models. The method includes obtaining, by one or more computing devices, a plurality of regions based at least in part on temporal availability of user devices; selecting a plurality of available user devices within a region; and providing a current version of a machine-learned model associated with the region to the plurality of selected user devices within the region. The method includes obtaining, from the plurality of selected user devices, updated machine-learned model data generated by the plurality of selected user devices through training of the current version of the machine-learned model associated with the region using data local to each of the plurality of selected user devices and generating an updated machine-learned model associated with the region based on the updated machine-learned model data.
    Type: Grant
    Filed: December 6, 2017
    Date of Patent: November 1, 2022
    Assignee: GOOGLE LLC
    Inventor: Keith Bonawitz
  • Publication number: 20210357986
    Abstract: Systems and methods are shown for providing private local sponsored content selection and improving intelligence models through distribution among mobile devices. This allows greater data gathering capabilities through the use of the sensors of the mobile devices as well as data stored on data storage components of the mobile devices to create predicted models while offering better opportunities to preserve privacy. Locally stored profiles comprising machine intelligence models may also be used to determine the relevance of the data gathered and in improving an aggregated model for identifying the relevance of data and the selection of sponsored content items. Distributed optimization is used in conjunction with privacy techniques to create the improved machine intelligence models. Publishers may also benefit from the improved privacy by protecting the statistics of type or volume of sponsored content items shown with publisher content.
    Type: Application
    Filed: July 29, 2021
    Publication date: November 18, 2021
    Applicant: Google LLC
    Inventors: Keith Bonawitz, Daniel Ramage, David Petrou
  • Patent number: 11087362
    Abstract: Systems and methods are shown for providing private local sponsored content selection and improving intelligence models through distribution among mobile devices. This allows greater data gathering capabilities through the use of the sensors of the mobile devices as well as data stored on data storage components of the mobile devices to create predicted models while offering better opportunities to preserve privacy. Locally stored profiles comprising machine intelligence models may also be used to determine the relevance of the data gathered and in improving an aggregated model for identifying the relevance of data and the selection of sponsored content items. Distributed optimization is used in conjunction with privacy techniques to create the improved machine intelligence models. Publishers may also benefit from the improved privacy by protecting the statistics of type or volume of sponsored content items shown with publisher content.
    Type: Grant
    Filed: November 25, 2019
    Date of Patent: August 10, 2021
    Assignee: GOOGLE LLC
    Inventors: Keith Bonawitz, Daniel Ramage, David Petrou
  • Publication number: 20200401946
    Abstract: The present disclosure provides systems and methods for the management and/or evaluation of machine-learned models based on locally logged data. In one example, a user computing device can obtain a machine-learned model (e.g., from a server computing device) and can evaluate at least one performance metric for the machine-learned model. In particular, the at least one performance metric for the machine-learned model can be evaluated relative to data that is stored locally at the user computing device. The user computing device and/or the server computing device can determine whether to activate the machine-learned model on the user computing device based at least in part on the at least one performance metric. In another example, the user computing device can evaluate a plurality of machine-learned models against locally stored data. At least one of the models can be selected based on the evaluated performance metrics.
    Type: Application
    Filed: September 8, 2020
    Publication date: December 24, 2020
    Inventors: Keith Bonawitz, Daniel Ramage
  • Patent number: 10769549
    Abstract: The present disclosure provides systems and methods for the management and/or evaluation of machine-learned models based on locally logged data. In one example, a user computing device can obtain a machine-learned model (e.g., from a server computing device) and can evaluate at least one performance metric for the machine-learned model. In particular, the at least one performance metric for the machine-learned model can be evaluated relative to data that is stored locally at the user computing device. The user computing device and/or the server computing device can determine whether to activate the machine-learned model on the user computing device based at least in part on the at least one performance metric. In another example, the user computing device can evaluate a plurality of machine-learned models against locally stored data. At least one of the models can be selected based on the evaluated performance metrics.
    Type: Grant
    Filed: November 21, 2016
    Date of Patent: September 8, 2020
    Assignee: Google LLC
    Inventors: Keith Bonawitz, Daniel Ramage
  • Publication number: 20200098011
    Abstract: Systems and methods are shown for providing private local sponsored content selection and improving intelligence models through distribution among mobile devices. This allows greater data gathering capabilities through the use of the sensors of the mobile devices as well as data stored on data storage components of the mobile devices to create predicted models while offering better opportunities to preserve privacy. Locally stored profiles comprising machine intelligence models may also be used to determine the relevance of the data gathered and in improving an aggregated model for identifying the relevance of data and the selection of sponsored content items. Distributed optimization is used in conjunction with privacy techniques to create the improved machine intelligence models. Publishers may also benefit from the improved privacy by protecting the statistics of type or volume of sponsored content items shown with publisher content.
    Type: Application
    Filed: November 25, 2019
    Publication date: March 26, 2020
    Applicant: Google LLC
    Inventors: Keith Bonawitz, Daniel Ramage, David Petrou
  • Patent number: 10504154
    Abstract: Systems and methods are shown for providing private local sponsored content selection and improving intelligence models through distribution among mobile devices. This allows greater data gathering capabilities through the use of the sensors of the mobile devices as well as data stored on data storage components of the mobile devices to create predicted models while offering better opportunities to preserve privacy. Locally stored profiles comprising machine intelligence models may also be used to determine the relevance of the data gathered and in improving an aggregated model for identifying the relevance of data and the selection of sponsored content items. Distributed optimization is used in conjunction with privacy techniques to create the improved machine intelligence models. Publishers may also benefit from the improved privacy by protecting the statistics of type or volume of sponsored content items shown with publisher content.
    Type: Grant
    Filed: September 20, 2016
    Date of Patent: December 10, 2019
    Assignee: Google LLC
    Inventors: Keith Bonawitz, Daniel Ramage, David Petrou
  • Patent number: 10324161
    Abstract: Disclosed herein are embodiments of a balloon-based positioning system and method. In one example embodiment, a system includes a group of at least three balloons deployed in the stratosphere and a control system configured for: determining a first set of spatial relationships relating to the group; determining a second set of spatial relationships relating to at least a portion of the group and to a reference point; determining a position of the reference point relative to the earth; using the determined first set, the determined second set, and the determined position of the reference point relative to the earth as a basis for determining a position of a target balloon in the group relative to the earth; and transmitting the determined position of the target balloon relative to the earth.
    Type: Grant
    Filed: November 6, 2017
    Date of Patent: June 18, 2019
    Assignee: Loon LLC
    Inventors: Keith Bonawitz, Richard Wayne DeVaul, Eric Teller, Joshua Weaver
  • Publication number: 20190171978
    Abstract: The present disclosure provides systems and methods for distributed training of machine learning models. In one example, a computer-implemented method is provided for training machine-learned models. The method includes obtaining, by one or more computing devices, a plurality of regions based at least in part on temporal availability of user devices; selecting a plurality of available user devices within a region; and providing a current version of a machine-learned model associated with the region to the plurality of selected user devices within the region. The method includes obtaining, from the plurality of selected user devices, updated machine-learned model data generated by the plurality of selected user devices through training of the current version of the machine-learned model associated with the region using data local to each of the plurality of selected user devices and generating an updated machine-learned model associated with the region based on the updated machine-learned model data.
    Type: Application
    Filed: December 6, 2017
    Publication date: June 6, 2019
    Inventor: Keith Bonawitz
  • Publication number: 20180144265
    Abstract: The present disclosure provides systems and methods for the management and/or evaluation of machine-learned models based on locally logged data. In one example, a user computing device can obtain a machine-learned model (e.g., from a server computing device) and can evaluate at least one performance metric for the machine-learned model. In particular, the at least one performance metric for the machine-learned model can be evaluated relative to data that is stored locally at the user computing device. The user computing device and/or the server computing device can determine whether to activate the machine-learned model on the user computing device based at least in part on the at least one performance metric. In another example, the user computing device can evaluate a plurality of machine-learned models against locally stored data. At least one of the models can be selected based on the evaluated performance metrics.
    Type: Application
    Filed: November 21, 2016
    Publication date: May 24, 2018
    Inventors: Keith Bonawitz, Daniel Ramage
  • Publication number: 20180067188
    Abstract: Disclosed herein are embodiments of a balloon-based positioning system and method. In one example embodiment, a system includes a group of at least three balloons deployed in the stratosphere and a control system configured for: determining a first set of spatial relationships relating to the group; determining a second set of spatial relationships relating to at least a portion of the group and to a reference point; determining a position of the reference point relative to the earth; using the determined first set, the determined second set, and the determined position of the reference point relative to the earth as a basis for determining a position of a target balloon in the group relative to the earth; and transmitting the determined position of the target balloon relative to the earth.
    Type: Application
    Filed: November 6, 2017
    Publication date: March 8, 2018
    Inventors: Keith Bonawitz, Richard Wayne DeVaul, Eric Teller, Joshua Weaver
  • Patent number: 9829561
    Abstract: Disclosed herein are embodiments of a balloon-based positioning system and method. In one example embodiment, a system includes a group of at least three balloons deployed in the stratosphere and a control system configured for: determining a first set of spatial relationships relating to the group; determining a second set of spatial relationships relating to at least a portion of the group and to a reference point; determining a position of the reference point relative to the earth; using the determined first set, the determined second set, and the determined position of the reference point relative to the earth as a basis for determining a position of a target balloon in the group relative to the earth; and transmitting the determined position of the target balloon relative to the earth.
    Type: Grant
    Filed: September 4, 2013
    Date of Patent: November 28, 2017
    Assignee: X Development LLC
    Inventors: Keith Bonawitz, Richard Wayne Devaul, Eric Teller, Joshua Weaver
  • Patent number: 9514269
    Abstract: Example methods and systems for determining failure modes of balloons within a balloon network are described. One example method includes: (a) determining at least one cohort balloon of a first balloon, where the first balloon is operating as part of a balloon network and where each cohort balloon shares at least one property with the first balloon, (b) determining at least one expected failure mode based at least in part on at least one failure of at least one cohort balloon, (c) determining a predicted failure mode of the first balloon based at least in part on the at least one expected failure mode, and (d) causing the first balloon to operate within the balloon network based at least in part on the predicted failure mode of the first balloon.
    Type: Grant
    Filed: July 17, 2013
    Date of Patent: December 6, 2016
    Assignee: X Development LLC
    Inventors: Keith Bonawitz, Joshua Weaver, Richard DeVaul
  • Patent number: 9424752
    Abstract: Example methods and systems for performing fleet planning based on coarse estimates of regions is provided. A method may include receiving information indicative of a sequence of coverage requirements for a region over a period of time. For one or more time intervals of the period of time, the method may include dividing the region over which vehicles of the plurality of vehicles may traverse into a plurality of sub-regions such that for each subsequent time interval a size of a given sub-region increases. The method includes at each of the one or more time intervals of the period of time, determining vehicles of the plurality of vehicles that can reach a given landmark in a given sub-region by an end of the one or more time intervals, and based on the sequence of coverage requirements, generating a fleet plan for the time intervals based on the determined vehicles.
    Type: Grant
    Filed: December 26, 2012
    Date of Patent: August 23, 2016
    Assignee: Google Inc.
    Inventor: Keith Bonawitz
  • Publication number: 20160173324
    Abstract: Example methods and systems for assigning tasks to balloons within a balloon network are described. One example system includes a first sub-fleet of balloons assigned a first set of one or more tasks within a balloon network, a second sub-fleet of balloons assigned a second set of one or more tasks within the balloon network, and a control system configured to determine that a first balloon in the first sub-fleet of balloons initially has a predicted failure mode that corresponds to the first set of tasks, subsequently determine that the first balloon has a predicted failure mode that corresponds to the second set of tasks, and reassign the first balloon from the first sub-fleet of balloons to the second sub-fleet of balloons.
    Type: Application
    Filed: February 19, 2016
    Publication date: June 16, 2016
    Inventors: Keith Bonawitz, Joshua Weaver, Richard DeVaul
  • Patent number: 9319905
    Abstract: Example methods and systems for assigning tasks to balloons within a balloon network are described. One example system includes a first sub-fleet of balloons assigned a first set of one or more tasks within a balloon network, a second sub-fleet of balloons assigned a second set of one or more tasks within the balloon network, and a control system configured to determine that a first balloon in the first sub-fleet of balloons initially has a predicted failure mode that corresponds to the first set of tasks, subsequently determine that the first balloon has a predicted failure mode that corresponds to the second set of tasks, and reassign the first balloon from the first sub-fleet of balloons to the second sub-fleet of balloons.
    Type: Grant
    Filed: August 30, 2013
    Date of Patent: April 19, 2016
    Assignee: Google Inc.
    Inventors: Keith Bonawitz, Joshua Weaver, Richard DeVaul
  • Publication number: 20150063159
    Abstract: Example methods and systems for assigning tasks to balloons within a balloon network are described. One example system includes a first sub-fleet of balloons assigned a first set of one or more tasks within a balloon network, a second sub-fleet of balloons assigned a second set of one or more tasks within the balloon network, and a control system configured to determine that a first balloon in the first sub-fleet of balloons initially has a predicted failure mode that corresponds to the first set of tasks, subsequently determine that the first balloon has a predicted failure mode that corresponds to the second set of tasks, and reassign the first balloon from the first sub-fleet of balloons to the second sub-fleet of balloons.
    Type: Application
    Filed: August 30, 2013
    Publication date: March 5, 2015
    Applicant: Google Inc.
    Inventors: Keith Bonawitz, Joshua Weaver, Richard DeVaul
  • Publication number: 20150061937
    Abstract: Disclosed herein are embodiments of a balloon-based positioning system and method. In one example embodiment, a system includes a group of at least three balloons deployed in the stratosphere and a control system configured for: determining a first set of spatial relationships relating to the group; determining a second set of spatial relationships relating to at least a portion of the group and to a reference point; determining a position of the reference point relative to the earth; using the determined first set, the determined second set, and the determined position of the reference point relative to the earth as a basis for determining a position of a target balloon in the group relative to the earth; and transmitting the determined position of the target balloon relative to the earth.
    Type: Application
    Filed: September 4, 2013
    Publication date: March 5, 2015
    Applicant: Google Inc.
    Inventors: Keith BONAWITZ, Richard Wayne DEVAUL, Eric TELLER, Joshua WEAVER
  • Patent number: 8775358
    Abstract: A method, apparatus and computer program product for performing probabilistic inference and providing related solution methods is presented. At least one state space (SS) is obtained for variables of interest relating to a problem of interest. None or more densities (D) defining pure functions over locations in the at least one SS are also obtained as is none or more kernels (K) defining a stochastic walk through the at least one SS. A virtual machine executes a stochastic walk through the state space to produce a solution for a problem of interest.
    Type: Grant
    Filed: November 24, 2008
    Date of Patent: July 8, 2014
    Assignee: Massachusetts Institute of Technology
    Inventors: Keith Bonawitz, Vikash Mansinghka
  • Publication number: 20090144218
    Abstract: A method, apparatus and computer program product for performing probabilistic inference and providing related solution methods is presented. At least one state space (SS) is obtained for variables of interest relating to a problem of interest. None or more densities (D) defining pure functions over locations in the at least one SS are also obtained as is none or more kernels (K) defining a stochastic walk through the at least one SS. A virtual machine executes a stochastic walk through the state space to produce a solution for a problem of interest.
    Type: Application
    Filed: November 24, 2008
    Publication date: June 4, 2009
    Inventors: Keith Bonawitz, Vikash Mansinghka