Artificial intelligence modifying federated learning models
Data verification in federate learning is faster and simpler. As artificial intelligence grows in usage, data verification is needed to prove custody and/or control. Electronic data representing an original version of training data may be hashed to generate one or more digital signatures. The digital signatures may then be incorporated into one or more blockchains for historical documentation. Any auditor may then quickly verify and/or reproduce the training data using the digital signatures. For example, a current version of the training data may be hashed and compared to the digital signatures generated from the current version of the training data. If the digital signatures match, then the training data has not changed since its creation. However, if the digital signatures do not match, then the training data has changed since its creation. The auditor may thus flag the training data for additional investigation and scrutiny.
1. A method executed by a server modifying a federated learning model based on a plurality of local updates reported by a plurality of client devices, the method comprising:
receiving, by the server, the plurality of local updates reported via a communications network by the plurality of client devices executing the federated learning model;
dedicating, by the server, a blockchain to only the federated learning model;
recording, by the server, the plurality of local updates reported by the plurality of client devices to the blockchain dedicated only to the federated learning model;
identifying a block of data chained within the blockchain dedicated to only a particular device of the plurality of client devices;
receiving a geographic location reported by the particular device; and
recording the geographic location to the block of data dedicated to only the particular device.
2. The method of claim 1 , wherein the local update reported by the particular device is recorded to the block of data dedicated to only the particular device.
3. The method of claim 1 , further comprising receiving a software identifier that uniquely identifies the federated learning model.
4. The method of claim 3 , further comprising recording the software identifier to the blockchain dedicated only to the federated learning model.
5. A server modifying a federated learning model based on a plurality of local updates reported by a plurality of client devices, comprising:
a hardware processor; and
a memory device storing instructions that when executed by the hardware processor perform operations, the operations comprising:
receiving the plurality of local updates reported via communications networks by the plurality of client devices executing the federated learning model;
dedicating a blockchain to only the federated learning model;
recording the plurality of local updates reported by the plurality of client devices to the blockchain dedicated only to the federated learning model;
identifying a block of data chained within the blockchain dedicated to only a particular device of the plurality of client devices;
receiving a geographic location reported by the particular device; and
recording the geographic location to the block of data dedicated to only the particular device.
6. The server of claim 5 , wherein the local update reported by the particular device is recorded to the block of data dedicated to only the particular device.
7. The server of claim 5 , wherein the operations further comprise receiving a software identifier that uniquely identifies the federated learning model.
8. The server of claim 7 , wherein the operations further comprise recording the software identifier to the blockchain dedicated only to the federated learning model.
9. A memory device storing instructions that when executed by a hardware processor perform operations, the operations comprising:
receiving a plurality of bit strings representing a plurality of local updates reported via communications networks by a plurality of client devices executing a federated learning model;
generating a plurality of shorter bit strings representing the plurality of local updates by hashing the plurality of bit strings;
dedicating a blockchain to only the federated learning model;
recording the plurality of shorter bit strings representing the plurality of local updates reported by the plurality of client devices to the blockchain dedicated only to the federated learning model;
identifying a block of data chained within the blockchain dedicated to only a particular device of the plurality of client devices;
receiving a geographic location reported by the particular device; and
recording the geographic location to the block of data dedicated to only the particular device.
10. The memory device of claim 9 , wherein the shorter bit string representing the local update_reported by the particular_device is recorded to the block of data dedicated to only the particular device.
11. The memory device of claim 9 , wherein the operations further comprise:
receiving a software identifier that uniquely identifies the federated learning model; and
recording the software identifier to the blockchain dedicated only to the federated learning model.