Sachin P Kamat

Title: Sachin P. Kamat: Inventor in Federated Learning Privacy

Introduction

Sachin P. Kamat is an inventor based in Bengaluru, India, and works at Intel Corporation. His patent work addresses privacy protection in federated learning, where models can be trained across distributed systems without directly sharing private data.

Latest Patents

Kamat is associated with 1 patent: “Mitigating Private Data Leakage in a Federated Learning System.” The invention describes receiving a model trained by a federated server using public data and global model parameters, then training it with locally shareable data to generate local model parameters. It further describes obscuring those parameters and sending them to the server. The model can then be adapted to classify private data and trained on that data using the local parameters, which are updated in the process.

Career Highlights

Kamat’s listed invention focuses on reducing the risk of private-data leakage while enabling collaborative model training. The approach distinguishes between shareable and private local data and uses parameter obscuring as part of its privacy-protection process.

Collaborations

Kamat’s listed coworkers include Jerome Anand and Anil Kumar Bangalore Rajashekar. His work is associated with Intel Corporation.

Conclusion

Sachin P. Kamat’s patent contribution centers on protecting private data during federated learning through local training, parameter obscuring, and model updates.