# US 2020/0394518 A1 · Method for collaborative learning of an artificial neural network without disclosing training data

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2020/0394518 A1
- Document ID: US20200394518A1
- Status: Published application
- Kind: A1
- Application number: 16/899,056
- Filed: 2020-06-11
- Published: 2020-12-17
- Earliest claimed: 2019-06-12
- Assignees: Commissariat A L'energie Atomique Et Aux Energies Alternatives
- Applicants: Commissariat A L'energie Atomique Et Aux Energies Alternatives
- Inventors: Renaud SIRDEY; Sergiu Carpov
- Web page: https://patentgrep.com/patent/US20200394518A1

## CPC

- G06N 3/063: using electronic means
- G06F 18/211: Selection of the most significant subset of features
- G06N 3/045: Combinations of networks
- G06N 3/08: Learning methods
- G06N 3/084: Backpropagation, e.g. using gradient descent
- G06N 3/09: Supervised learning
- G06N 3/098: Distributed learning, e.g. federated learning
- H04L 9/008: {involving homomorphic encryption}
- H04L 9/085: {Secret sharing or secret splitting, e.g. threshold schemes}
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)

## Abstract

A method for federative learning of an artificial neural network model on a plurality of sets of training data. The learning method used a plurality of data suppliers each having a distinct set of training data and a symmetric key, and an aggregation platform aggregating partial models trained on a sub-plurality of these sets, in each iteration. In each iteration, the platform selects a sub-plurality of data suppliers and supplies the parameters of the model to them, in homomorphically encrypted form. Each training data supplier decrypts these parameters, trains the model on his own data set and returns the parameters of the partial model thus obtained, encrypted by his symmetric key, to the platform. The aggregation platform then transcrypts these partial models in the homomorphic domain, and combines them in the homomorphic domain to obtain a global model.

## Claims

### Claim 1

A method for federative learning of a model of an artificial neural network, the model being characterised by parameters giving synaptic coefficients of said network, the method making use of a plurality of N data suppliers (DS1, . . . , DSN) each having a distinct set of learning data (S1, . . . , SN) and an aggregation platform (FP), said method beginning with an initialisation step to initialise the parameters of the model, and proceeding with a plurality of successive iterations, each iteration updating the parameters of a current model of the neural network, wherein:
in the initialisation step, the data suppliers share the private key and the public key of a homomorphic cryptosystem (HE.sk,HE.pk), each data supplier also generating a secret stream encryption key, Ksymn, and transmitting the public key of the homomorphic cryptosystem encrypted using said secret key, Enc(Ksymn;HE.pk) to the aggregation platform;
and wherein each iteration t,
the aggregation platform selects a sub-plurality M≤N of data suppliers and transmits the parameters of the current model, homomorphically encrypted, Enc(wt,HE.pk), to the M data suppliers;
each selected data supplier decrypts the parameters thus encrypted using the private key of the homomorphic cryptosystem, trains the current model on its learning data set to obtain a partial model;
each selected data supplier, DSn, encrypts) the parameters of the partial model using his secret key, Ksymn, and transmits the parameters of the partial model thus encrypted, Enc(wnt,Ksymn), to the aggregation platform;
the aggregation platform transcrypts parameters of the partial models to the different selected data suppliers, to obtain these same parameters encrypted in the homomorphic domain,
the aggregation platform combines, in the homomorphic domain, parameters of the partial models obtained by the different data suppliers, to obtain the parameters of a new current model encrypted in the homomorphic domain;
and, when a stop criterion is satisfied, each data supplier decrypts the parameters of the last current model using the private key of the homomorphic cryptosystem, to obtain a final model, trained on the reunion of said learning data sets.

### Claim 2 (depends on 1)

The method of federative learning of a model of an artificial neural network according to claim 1, wherein the aggregation platform averages the parameters of each of the partial models thus transcrypted, in the homomorphic domain.

### Claim 3 (depends on 1)

The method of federative learning of a model of an artificial neural network according to claim 1, wherein the secret keys of the different data suppliers are identical, the common secret key having been shared between the different data suppliers using a multipartite key exchange protocol.

### Claim 4 (depends on 3)

The method of federative learning of a model of an artificial neural network according to claim 3, wherein the platform averages the different parameters of the partial models by batch processing on a homomorphic encrypted value representing a composite of the different parameters of the partial models.

### Claim 5 (depends on 3)

The method of federative learning of a model of an artificial neural network according to claim 3, wherein the platform averages the parameters of the same rank of the partial models, by batch processing on a homomorphic encrypted value representing a composite of the parameters of said rank of partial models, using batch processing on a homomorphic encrypted value representing a composite of the parameters of said rank of partial models from the different data suppliers.