Patents by Inventor Abhradeep Guha Thakurta
Abhradeep Guha Thakurta 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).
-
Publication number: 20240095594Abstract: A method includes training a first differentially private (DP) model using a private training set, the private training set including a plurality of training samples, the first DP model satisfying a differential privacy budget, the differential privacy budget defining an amount of information about individual training samples of the private training set that may be revealed by the first DP model. The method also includes, while training the first DP model, generating a plurality of intermediate checkpoints, each intermediate checkpoint of the plurality of intermediate checkpoints representing a different intermediate state of the first DP model, each of the intermediate checkpoints satisfying the same differential privacy budget. The method further includes determining an aggregate of the first DP model and the plurality of intermediate checkpoints, and determining, using the aggregate, a second DP model, the second DP model satisfying the same differential privacy budget.Type: ApplicationFiled: August 31, 2023Publication date: March 21, 2024Applicant: Google LLCInventors: Om Dipakbhai Thakkar, Arun Ganesh, Virat Vishnu Shejwalkar, Abhradeep Guha Thakurta, Rajiv Mathews
-
Publication number: 20240054391Abstract: Computer-implemented systems and methods for training a decentralized model for making a personalized recommendation.Type: ApplicationFiled: April 5, 2022Publication date: February 15, 2024Inventors: Abhradeep Guha Thakurta, Li Zhang, Prateek Jain, Shuang Song, Steffen Rendle, Steve Shaw-Tang Chien, Walid Krichene, Yarong Mu
-
Publication number: 20230223028Abstract: Techniques are disclosed that enable training a global model using gradients provided to a remote system by a set of client devices during a reporting window, where each client device randomly determines a reporting time in the reporting window to provide the gradient to the remote system. Various implementations include each client device determining a corresponding gradient by processing data using a local model stored locally at the client device, where the local model corresponds to the global model.Type: ApplicationFiled: October 16, 2020Publication date: July 13, 2023Inventors: Om Thakkar, Abhradeep Guha Thakurta, Peter Kairouz, Borja de Balle Pigem, Brendan McMahan
-
Publication number: 20230103911Abstract: A method include obtaining a set of differentially private (DP) gradients each generated based on processing corresponding private data, and obtaining a set of public gradients each generated based on processing corresponding public data. The method also includes applying mirror descent to the set of public gradients to learn a geometry for the set of DP gradients, and reshaping the set of DP gradients based on the learned geometry. The method further includes training a machine learning model based on the reshaped set of DP gradients.Type: ApplicationFiled: October 4, 2022Publication date: April 6, 2023Applicant: Google LLCInventors: Om Dipakbhai Thakkar, Ehsan Amid, Arun Ganesh, Rajiv Mathews, Swaroop Ramaswamy, Shuang Song, Thomas Steinke, Vinith Suriyakumar, Abhradeep Guha Thakurta
-
Patent number: 11042664Abstract: One embodiment provides a system that implements a 1-bit protocol for differential privacy for a set of client devices that transmit information to a server. Implementations may leverage specialized instruction sets or engines built into the hardware or firmware of a client device to improve the efficiency of the protocol. For example, a client device may utilize these cryptographic functions to randomize information sent to the server. In one embodiment, the client device may use cryptographic functions such as hashes including SHA or block ciphers including AES to provide an efficient mechanism for implementing differential privacy.Type: GrantFiled: January 17, 2020Date of Patent: June 22, 2021Assignee: Apple Inc.Inventors: Yannick L. Sierra, Abhradeep Guha Thakurta, Umesh S. Vaishampayan, John C. Hurley, Keaton F. Mowery, Michael Brouwer
-
Patent number: 11023594Abstract: Technologies are disclosed for computing heavy hitter histograms using locally private randomization. Under this strategy, “agents” can each hold a “type” derived from a large dictionary. By performing an algorithm, an estimate of the distribution of data can be obtained. Two algorithms implement embodiments for performing methods involving differential privacy for one or more users, and usually are run in the local model. This means that information is collected from the agents with added noise to hide the agents' individual contributions to the histogram. The result is an accurate enough estimate of the histogram for commercial or other applications relating to the data collection of one or more agents. Specifically, the proposed algorithms improve on the performance (measured in computation and memory requirements at the server and the agent, as well as communication volume) of previously solutions.Type: GrantFiled: May 22, 2018Date of Patent: June 1, 2021Assignee: Georgetown UniversityInventors: Yaacov Nissim Kobliner, Uri Stemmer, Raef Bahi Youssef Bassily, Abhradeep Guha Thakurta
-
Publication number: 20200257816Abstract: One embodiment provides a system that implements a 1-bit protocol for differential privacy for a set of client devices that transmit information to a server. Implementations may leverage specialized instruction sets or engines built into the hardware or firmware of a client device to improve the efficiency of the protocol. For example, a client device may utilize these cryptographic functions to randomize information sent to the server. In one embodiment, the client device may use cryptographic functions such as hashes including SHA or block ciphers including AES to provide an efficient mechanism for implementing differential privacy.Type: ApplicationFiled: January 17, 2020Publication date: August 13, 2020Inventors: Yannick L. Sierra, Abhradeep Guha Thakurta, Umesh S. Vaishampayan, John C. Hurley, Keaton F. Mowery, Michael Brouwer
-
Patent number: 10701042Abstract: Systems and methods are disclosed for a server learning new words generated by user client devices in a crowdsourced manner while maintaining local differential privacy of client devices. A client device can determine that a word typed on the client device is a new word that is not contained in a dictionary or asset catalog on the client device. New words can be grouped in classifications such as entertainment, health, finance, etc. A differential privacy system on the client device can comprise a privacy budget for each classification of new words. If there is privacy budget available for the classification, then one or more new terms in a classification can be sent to new term learning server, and the privacy budget for the classification reduced. The privacy budget can be periodically replenished.Type: GrantFiled: October 12, 2018Date of Patent: June 30, 2020Assignee: Apple Inc.Inventors: Abhradeep Guha Thakurta, Andrew H. Vyrros, Umesh S. Vaishampayan, Gaurav Kapoor, Julien Freudiger, Vivek Rangarajan Sridhar, Doug Davidson
-
Patent number: 10552631Abstract: One embodiment provides a system that implements a 1-bit protocol for differential privacy for a set of client devices that transmit information to a server. Implementations may leverage specialized instruction sets or engines built into the hardware or firmware of a client device to improve the efficiency of the protocol. For example, a client device may utilize these cryptographic functions to randomize information sent to the server. In one embodiment, the client device may use cryptographic functions such as hashes including SHA or block ciphers including AES to provide an efficient mechanism for implementing differential privacy.Type: GrantFiled: March 8, 2019Date of Patent: February 4, 2020Assignee: Apple Inc.Inventors: Yannick L. Sierra, Abhradeep Guha Thakurta, Umesh S. Vaishampayan, John C. Hurley, Keaton F. Mowery, Michael Brouwer
-
Patent number: 10454962Abstract: Systems and methods are disclosed for generating term frequencies of known terms based on crowdsourced differentially private sketches of the known terms. An asset catalog can be updated with new frequency counts for known terms based on the crowdsourced differentially private sketches. Known terms can have a classification. A client device can maintain a privacy budget for each classification of known terms. Classifications can include emojis, deep links, locations, finance terms, and health terms, etc. A privacy budget ensures that a client does not transmit too much information to a term frequency server, thereby compromising the privacy of the client device.Type: GrantFiled: October 12, 2018Date of Patent: October 22, 2019Assignee: Apple Inc.Inventors: Abhradeep Guha Thakurta, Andrew H. Vyrros, Umesh S. Vaishampayan, Gaurav Kapoor, Julien Freudiger, Vipul Ved Prakash, Arnaud Legendre, Steven Duplinsky
-
Publication number: 20190205561Abstract: One embodiment provides a system that implements a 1-bit protocol for differential privacy for a set of client devices that transmit information to a server. Implementations may leverage specialized instruction sets or engines built into the hardware or firmware of a client device to improve the efficiency of the protocol. For example, a client device may utilize these cryptographic functions to randomize information sent to the server. In one embodiment, the client device may use cryptographic functions such as hashes including SHA or block ciphers including AES to provide an efficient mechanism for implementing differential privacy.Type: ApplicationFiled: March 8, 2019Publication date: July 4, 2019Inventors: Yannick L. Sierra, Abhradeep Guha Thakurta, Umesh S. Vaishampayan, John C. Hurley, Keaton F. Mowery, Michael Brouwer
-
Publication number: 20190097978Abstract: Systems and methods are disclosed for a server learning new words generated by user client devices in a crowdsourced manner while maintaining local differential privacy of client devices. A client device can determine that a word typed on the client device is a new word that is not contained in a dictionary or asset catalog on the client device. New words can be grouped in classifications such as entertainment, health, finance, etc. A differential privacy system on the client device can comprise a privacy budget for each classification of new words. If there is privacy budget available for the classification, then one or more new terms in a classification can be sent to new term learning server, and the privacy budget for the classification reduced. The privacy budget can be periodically replenished.Type: ApplicationFiled: October 12, 2018Publication date: March 28, 2019Inventors: Abhradeep Guha Thakurta, Andrew H. Vyrros, Umesh S. Vaishampayan, Gaurav Kapoor, Julien Freudiger, Vivek Rangarajan Sridhar, Doug Davidson
-
Patent number: 10229282Abstract: The system described may implement a 1-bit protocol for differential privacy for a set of client devices that transmit information to a server. Implementations of the system may leverage specialized instruction sets or engines built into the hardware or firmware of a client device to improve the efficiency of the protocol. For example, a client device may utilize these cryptographic functions to randomize information sent to the server. In one embodiment, the client device may use cryptographic functions such as hashes including SHA or block ciphers including AES. Accordingly, the system provides an efficient mechanism for implementing differential privacy.Type: GrantFiled: September 23, 2016Date of Patent: March 12, 2019Assignee: Apple Inc.Inventors: Yannick L. Sierra, Abhradeep Guha Thakurta, Umesh S. Vaishampayan, John C. Hurley, Keaton F. Mowery, Michael Brouwer
-
Publication number: 20190068628Abstract: Systems and methods are disclosed for generating term frequencies of known terms based on crowdsourced differentially private sketches of the known terms. An asset catalog can be updated with new frequency counts for known terms based on the crowdsourced differentially private sketches. Known terms can have a classification. A client device can maintain a privacy budget for each classification of known terms. Classifications can include emojis, deep links, locations, finance terms, and health terms, etc. A privacy budget ensures that a client does not transmit too much information to a term frequency server, thereby compromising the privacy of the client device.Type: ApplicationFiled: October 12, 2018Publication date: February 28, 2019Inventors: Abhradeep Guha Thakurta, Andrew H. Vyrros, Umesh S. Vaishampayan, Gaurav Kapoor, Julien Freudiger, Vipul Ved Prakash, Arnaud Legendre, Steven Duplinsky
-
Patent number: 10154054Abstract: Systems and methods are disclosed for generating term frequencies of known terms based on crowdsourced differentially private sketches of the known terms. An asset catalog can be updated with new frequency counts for known terms based on the crowdsourced differentially private sketches. Known terms can have a classification. A client device can maintain a privacy budget for each classification of known terms. Classifications can include emojis, deep links, locations, finance terms, and health terms, etc. A privacy budget ensures that a client does not transmit too much information to a term frequency server, thereby compromising the privacy of the client device.Type: GrantFiled: June 30, 2017Date of Patent: December 11, 2018Assignee: Apple Inc.Inventors: Abhradeep Guha Thakurta, Andrew H. Vyrros, Umesh S. Vaishampayan, Gaurav Kapoor, Julien Freudinger, Vipul Ved Prakash, Arnaud Legendre, Steven Duplinsky
-
Publication number: 20180336357Abstract: Technologies are disclosed for computing heavy hitter histograms using locally private randomization. Under this strategy, “agents” can each hold a “type” derived from a large dictionary. By performing an algorithm, an estimate of the distribution of data can be obtained. Two algorithms implement embodiments for performing methods involving differential privacy for one or more users, and usually are run in the local model. This means that information is collected from the agents with added noise to hide the agents' individual contributions to the histogram. The result is an accurate enough estimate of the histogram for commercial or other applications relating to the data collection of one or more agents. Specifically, the proposed algorithms improve on the performance (measured in computation and memory requirements at the server and the agent, as well as communication volume) of previously solutions.Type: ApplicationFiled: May 22, 2018Publication date: November 22, 2018Applicants: Georgetown University, President and Fellows of Harvard College, The Regents of the University of CaliforniaInventors: Yaacov Nissim Kobliner, Uri Stemmer, Raef Bahi Youssef Bassily, Abhradeep Guha Thakurta
-
Patent number: 10133725Abstract: Systems and methods are disclosed for a server learning new words generated by user client devices in a crowdsourced manner while maintaining local differential privacy of client devices. A client device can determine that a word typed on the client device is a new word that is not contained in a dictionary or asset catalog on the client device. New words can be grouped in classifications such as entertainment, health, finance, etc. A differential privacy system on the client device can comprise a privacy budget for each classification of new words. If there is privacy budget available for the classification, then one or more new terms in a classification can be sent to new term learning server, and the privacy budget for the classification reduced. The privacy budget can be periodically replenished.Type: GrantFiled: April 3, 2017Date of Patent: November 20, 2018Assignee: Apple Inc.Inventors: Abhradeep Guha Thakurta, Andrew H. Vyrros, Umesh S. Vaishampayan, Gaurav Kapoor, Julien Freudiger, Vivek Rangarajan Sridhar, Doug Davidson
-
Patent number: 9894089Abstract: Systems and methods are disclosed for generating term frequencies of known terms based on crowdsourced differentially private sketches of the known terms. An asset catalog can be updated with new frequency counts for known terms based on the crowdsourced differentially private sketches. Known terms can have a classification. A client device can maintain a privacy budget for each classification of known terms. Classifications can include emojis, deep links, locations, finance terms, and health terms, etc. A privacy budget ensures that a client does not transmit too much information to a term frequency server, thereby compromising the privacy of the client device.Type: GrantFiled: June 19, 2017Date of Patent: February 13, 2018Assignee: Apple Inc.Inventors: Abhradeep Guha Thakurta, Andrew H. Vyrros, Umesh S. Vaishampayan, Gaurav Kapoor, Julien Freudinger, Vipul Ved Prakash, Arnaud Legendre, Steven Duplinsky
-
Publication number: 20180039619Abstract: Systems and methods are disclosed for a server learning new words generated by user client devices in a crowdsourced manner while maintaining local differential privacy of client devices. A client device can determine that a word typed on the client device is a new word that is not contained in a dictionary or asset catalog on the client device. New words can be grouped in classifications such as entertainment, health, finance, etc. A differential privacy system on the client device can comprise a privacy budget for each classification of new words. If there is privacy budget available for the classification, then one or more new terms in a classification can be sent to new term learning server, and the privacy budget for the classification reduced. The privacy budget can be periodically replenished.Type: ApplicationFiled: April 3, 2017Publication date: February 8, 2018Inventors: Abhradeep Guha Thakurta, Andrew H. Vyrros, Umesh S. Vaishampayan, Gaurav Kapoor, Julien Freudiger, Vivek Rangarajan Sridhar, Doug Davidson
-
Publication number: 20170359363Abstract: Systems and methods are disclosed for generating term frequencies of known terms based on crowdsourced differentially private sketches of the known terms. An asset catalog can be updated with new frequency counts for known terms based on the crowdsourced differentially private sketches. Known terms can have a classification. A client device can maintain a privacy budget for each classification of known terms. Classifications can include emojis, deep links, locations, finance terms, and health terms, etc. A privacy budget ensures that a client does not transmit too much information to a term frequency server, thereby compromising the privacy of the client device.Type: ApplicationFiled: June 19, 2017Publication date: December 14, 2017Inventors: Abhradeep Guha Thakurta, Andrew H. Vyrros, Umesh S. Vaishampayan, Gaurav Kapoor, Julien Freudinger, Vipul Ved Prakash, Arnaud Legendre, Steven Duplinsky