IP Library › Granted Patent US 12,608,619
Granted Patent B2
US 12,608,619 · App. 17/559,159 · Granted Apr 21, 2026

Superseded federated learning

Inventors: Ohad Arnon (Beit-Nir, IL); Dany Shapiro (Alfei Menashe, IL)
Assignee: Dell Products L.P.
G06N3/088G06N3/045
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12,608,619
App. No.
17/559,159
Granted
Apr 21, 2026
Kind
B2
Abstract

A method and system for implementing superseded federated learning. Superseded federated learning may entail a novel, performance-efficient federated learning technique designed to further decouple multiparty dependency on one another, as well as any third-parties, while collaborating in multiparty computations. Specifically, unlike any current federated learning methodology, superseded federated learning eliminates the complex and often inefficient coordination amongst parties, as well as removes third-party participation, during the classification or prediction inference phase of multiparty collaborations.

Claims (64)

1 . A method for inferring outcomes using superseded federated learning on a system, the method comprising:

identifying, on a node of a plurality of nodes of the system, a set of common features of data sets stored on different nodes of a plurality of nodes by implementing a private set intersection (PSI) between the plurality of nodes to ascertain an intersection of the data sets of the different nodes without exposing respective data sets of the different nodes to other nodes;

identifying a set of common sample identifiers (IDs) based on a common feature of the set of common features, wherein the PSI is applied to varying feature value vectors, representing the common feature of the set of common features, to identify the set of common sample IDs;

obtaining a local domain dataset mapped to the set of common sample IDs, wherein the local domain dataset is a portion of the local domain data and wherein the local domain data comprises data of the node;

initializing, after obtaining the local domain dataset, a local learning model by configuring a learning state with respective default values;

producing a trained local learning model from the local learning model by training on the local domain dataset, wherein training entails iterative adjustments to the learning state until an acceptable performance condition is reached;

initializing, after producing the trained local learning model, a local generative adversarial network (GAN), wherein the local GAN comprises a local generator model and a local discriminator model and wherein the local generator model and the local discriminator model are pit against each other to generate synthetic data;

producing a trained local GAN from the local GAN based on a local generator model training data set and a local discriminator model training data set, wherein the local generator model training set comprises the data of the node mapped to the set of common features and wherein the local discriminator model training data set comprises second data sampled from the local domain data;

extracting a trained local generator model from the trained local GAN;

sharing the trained local generator model and the local learning model with at least one other remote node of the plurality of nodes via a network, wherein the at least one other remote node is separate from the node;

receiving, after sharing and from the at least one other remote node of the system, one or more trained remote generator models and one or more trained remote learning models via the network;

after receiving the one or more trained remote generator models and the one or more trained remote learning models:

obtaining a new sample at the node of the system;

processing the new sample using the trained local learning model to obtain a local model outcome;

deriving a remote generator model input data set from the new sample;

processing the remote generator model input data set using the one or more trained remote generator models of the plurality of nodes to obtain at least one synthesized remote learning model input data set, wherein the at least one synthesized remote learning model input data set is created to simulate sample data from separate nodes of the plurality of nodes;

processing the at least one synthesized remote learning model input data set using the one or more trained remote learning models on the node to obtain at least one remote model outcome; and

aggregating the local model outcome and the at least one remote model outcome using a local ensemble learning model of the node to produce an aggregated outcome, wherein the local ensemble learning model comprises a machine learning model configured to aggregate the local model outcome and the at least one remote model outcome using an ensemble learning technique.

2 . The method of claim 1 , wherein the new sample comprises an ordered list of data values mapped, respectively, to a feature space relevant to a local domain.

3 . The method of claim 2 , wherein the feature space comprises a set of common features also exhibited in at least another feature space relevant to at least one remote domain.

4 . The method of claim 3 , wherein the remote generator model input data set comprises a subset of the new sample, wherein the subset of the new sample comprises a subset of the ordered list of data values mapped to the set of common features.

5 . The method of claim 3 , wherein the at least one synthesized remote learning model input data each comprises an ordered list of data values mapped, respectively, to the at least another feature space relevant to the at least one remote domain.

6 . The method of claim 1 , wherein the common feature reflects identification information uniquely identifying a set of entities, wherein local domain data pertinent to a local domain comprises data for the set of entities.

7 . The method of claim 6 , wherein the common feature further reflects identification information uniquely identifying at least another set of entities, wherein remote domain data pertinent to at least one remote domain comprises data for the at least another set of entities.

8 . The method of claim 6 , wherein the local learning model training data set comprises a subset of the local domain data, wherein the subset of the local domain data comprises a subset of samples mapped to the set of common sample IDs.

9 . The method of claim 1 , wherein the trained local generator model results from optimally configuring a local generator model using the local generator model training data set, wherein the trained local discriminator model results from optimally configuring a local discriminator model using the local discriminator model training data set.

10 . A non-transitory computer readable medium (CRM) comprising computer readable program code, which when executed by a computer processor, enables the computer processor to perform a method for inferring outcomes using superseded federated learning on a system, the method comprising:

identifying, on a node of a plurality of nodes of the system, a set of common features of data sets stored on different nodes of a plurality of nodes by implementing a private set intersection (PSI) between the plurality of nodes to ascertain an intersection of the data sets of the different nodes without exposing respective data sets of the different nodes to other nodes;

identifying a set of common sample identifiers (IDs) based on a common feature of the set of common features, wherein the PSI is applied to varying feature value vectors, representing the common feature of the set of common features, to identify the set of common sample IDs;

obtaining a local domain dataset mapped to the set of common sample IDs, wherein the local domain dataset is a portion of the local domain data and wherein the local domain data comprises data of the node;

initializing, after obtaining the local domain dataset, a local learning model by configuring a learning state with respective default values;

producing a trained local learning model from the local learning model by training on the local domain dataset, wherein training entails iterative adjustments to the learning state until an acceptable performance condition is reached;

initializing, after producing the trained local learning model, a local generative adversarial network (GAN), wherein the local GAN comprises a local generator model and a local discriminator model and wherein the local generator model and the local discriminator model are pit against each other to generate synthetic data;

producing a trained local GAN from the local GAN based on a local generator model training data set and a local discriminator model training data set, wherein the local generator model training set comprises the data of the node mapped to the set of common features and wherein the local discriminator model training data set comprises second data sampled from the local domain data;

extracting a trained local generator model from the trained local GAN;

sharing the trained local generator model and the local learning model with at least one other remote node of the plurality of nodes via a network, wherein the at least one other remote node is separate from the node;

receiving, after sharing and from the at least one other remote node of the system, one or more trained remote generator models and one or more trained remote learning models via the network;

after receiving the one or more trained remote generator models and the one or more trained remote learning models:

obtaining a new sample at the node of a system;

processing the new sample using the trained local learning model to obtain a local model outcome;

deriving a remote generator model input data set from the new sample;

processing the remote generator model input data set using the one or more trained remote generator models of the plurality of nodes to obtain at least one synthesized remote learning model input data set, wherein the at least one synthesized remote learning model input data set is created to simulate sample data from separate nodes of the plurality of nodes

processing the at least one synthesized remote learning model input data set using the one or more trained remote learning models on the node to obtain at least one remote model outcome; and

aggregating the local model outcome and the at least one remote model outcome using a local ensemble learning model of the node to produce an aggregated outcome, wherein the local ensemble learning model comprises a machine learning model configured to aggregate the local model outcome and the at least one remote model outcome using an ensemble learning technique.

11 . A system, comprising:

a plurality of nodes operatively connected to one another through a network, wherein each node of the plurality of nodes comprises a computer processor configured to perform a method for inferring outcomes using superseded federated learning, the method comprising:

identifying a set of common features of data sets stored on different nodes of the plurality of nodes by implementing a private set intersection (PSI) between the plurality of nodes to ascertain an intersection of the data sets of the different nodes without exposing respective data sets of the different nodes to other nodes;

identifying a set of common sample identifiers (IDs) based on a common feature of the set of common features, wherein the PSI is applied to varying feature value vectors, representing the common feature of the set of common features, to identify the set of common sample IDs;

obtaining a local domain dataset mapped to the set of common sample IDs, wherein the local domain dataset is a portion of the local domain data and wherein the local domain data comprises data of the node;

initializing, after obtaining the local domain dataset, a local learning model by configuring a learning state with respective default values;

producing a trained local learning model from the local learning model by training on the local domain dataset, wherein training entails iterative adjustments to the learning state until an acceptable performance condition is reached;

initializing, after producing the trained local learning model, a local generative adversarial network (GAN), wherein the local GAN comprises a local generator model and a local discriminator model and wherein the local generator model and the local discriminator model are pit against each other to generate synthetic data;

producing a trained local GAN from the local GAN based on a local generator model training data set and a local discriminator model training data set, wherein the local generator model training set comprises the data of the node mapped to the set of common features and wherein the local discriminator model training data set comprises second data sampled from the local domain data;

extracting a trained local generator model from the trained local GAN;

sharing the trained local generator model and the local learning model with at least one other remote node of the plurality of nodes via a network, wherein the at least one other remote node is separate from the node;

receiving, after sharing and from the at least one other remote node of the system, one or more trained remote generator models and one or more trained remote learning models via the network;

after receiving the one or more trained remote generator models and the one or more trained remote learning models;

obtaining a new sample;

processing the new sample using the trained local learning model to obtain a local model outcome;

deriving a remote generator model input data set from the new sample;

processing the remote generator model input data set using the one or more trained remote generator models of the plurality of nodes to obtain at least one synthesized remote learning model input data set, wherein

the at least one synthesized remote learning model input data set is created to simulate sample data from separate nodes of the plurality of nodes;

processing the at least one synthesized remote learning model input data set using the one or more trained remote learning models on the node to obtain at least one remote model outcome; and

aggregating the local model outcome and the at least one remote model outcome using a local ensemble learning model of the node to produce an aggregated outcome, wherein the local ensemble learning model comprises a machine learning model configured to aggregate the local model outcome and the at least one remote model outcome using an ensemble learning technique.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Jan 25, 2022
From: ARNON, OHAD; SHAPIRO, DANY
To: DELL PRODUCTS L.P.
Reel/Frame 058758/0762 →
Continuity (1)
Related Publication 20230196115A1 · Jun 22, 2023
References Cited (21)
US 20220058494A1 · Wu · 2022 [cited by examiner]
US 20220398343A1 · Ou · 2022 [cited by examiner]
CN 113468521A · 2021 [cited by examiner]
Mendes, Andre et al.; “Unified Multi-Domain Learning and Data Imputation using Adversarial Autoencoder”; Mar. 2020; arXiv; 1-8 (Year: 2020). [cited by examiner]
Lu, Linpeng et al.; “Multi-Party Private Set Intersection in Vertical Federated Learning”; 2020; IEEE 19th International Conference on Trust, Security, and Privacy in Computing and Communications; 707-714 (Year: 2020). [cited by examiner]
Ou, Wei et al.; “A Homomorphic-encryption-based Vertical Federated Learning Scheme for Rick Management”; 2020; Computer Science and Information Systems 17(3); 819-834 (Year: 2020). [cited by examiner]
Lu, Yunlong; “Blockchain and Federated Learning for Privacy-Preserved Data Sharing in Industrial IoT”; Jun. 2020; IEEE Transactions on Industrial Informatics vol. 15 No. 6; (Year: 2020). [cited by examiner]
Li, Rui et al.; “Generating Target Image-Label Pairs for Unsupervised Domain Adaptation”; Jul. 2020; IEEE Transactions on Image Processing vol. 29; 7997-8011 (Year: 2020). [cited by examiner]
Chang, Qi et al.; “Synthetic Learning: Learn From Distributed Asynchronized Discriminator GAN Without Sharing Medical Image Data”; Jun. 2020; arXiv; 1-15 (Year: 2020). [cited by examiner]
Website entitled “Building Your Own Federated Learning Algorithm”; Aug. 2021; TensorFlow; Retrieved from Internet: <https://www.tensorflow.org/federated/tutorials/building_your_own_federated_learning_algorithm> (Year: 2… [cited by examiner]
Goetz, Jack et al.; “Federated Learning via Synthetic Data”; Sep. 2020; arXiv; 1-12 (Year: 2020). [cited by examiner]
Angelou, Nick et al.; “Asymmetric Private Set Intersection with Applications to Contact Tracing and Private Vertical Federated Machine Learning”; Nov. 2020; arXiv; 1-10 (Year: 2020). [cited by examiner]
Xin, Bangzhou et al.; “Private FL-GAN: Differential Privacy Synthetic Data Generation Based on Federated Learning”; 2020; IEEE ICASSP 2020; 2927-2931 (Year: 2927). [cited by examiner]
Lin, Tao et al.; Ensemble Distillation for Robust Model Fusion in Federated Learning; Mar. 2021; arXiv; 1-26 (Year: 2021). [cited by examiner]
Bouacida, N. & Mohapatra, P. “Vulnerabilities in Federated Learning.” IEEE Access, vol. 9, 2021, pp. 63229-63249. (Mar. 16, 2021) (21 pages). [cited by applicant]
Jatain, D., Singh, V. & Daihiya, N. “A contemplative perspective on federated machine learning: Taxonomy, threats & vulnerability assessment and challenges.” Journal of King Saud University—Computer and Information Scie… [cited by applicant]
Kairouz, P., et al. “Advances and Open Problems in Federated Learning.” Foundations and Trends in Machine Learning, vol. 14, Issue 1-2, 2021, (122 pages). [cited by applicant]
Li, T., Sahu, A. K., Talwalkar, A. & Smith, V. “Federated Learning: Challenges, Methods, and Future Directions.” IEEE Signal Processing Magazine, vol. 37, 2020, (21 pages). [cited by applicant]
Ma, J., Naas, S. A., Sigg, S. & Lyu, X. “Privacy-preserving Federated Learning based on Multi-key Homomorphic Encryption.” ArXiv:2104.06824, 2021. (16 pages). [cited by applicant]
Ogburn, M., Turner, C. & Dahal, P. “Homomorphic Encryption.” Procedia Computer Science, vol. 20, 2013, pp. 502-509. (8 pages). [cited by applicant]
Yang, Q., Liu, Y., Chen, T. & Tong, Y. “Federated Machine Learning: Concept and Applications.” ACM Transactions on Intelligent Systems and Technology (TIST), vol. 10, Issue 2, 2019, pp. 1-19. (20 pages). [cited by applicant]