Pallika Haridas Kanani

Title: Innovations of Pallika Haridas Kanani

Introduction

Pallika Haridas Kanani is an accomplished inventor based in Westford, MA (US). He has made significant contributions to the field of machine learning and privacy preservation. With a total of five patents to his name, Kanani is recognized for his innovative approaches to federated learning and data privacy.

Latest Patents

One of his latest patents is titled "Subject Level Privacy Attack Analysis for Federated Learning." This invention allows for the analysis of inference attacks to determine the presence of a subject's data in a training set of a federated machine learning model. The process involves generating success measurements based on the performance of the selected inference attacks.

Another notable patent is "Privacy Preserving Collaborative Learning with Domain Adaptation." This invention introduces techniques for domain adaptation of a machine learning model while imposing differential privacy onto federated learning. The method involves client devices receiving coefficients from a server and training private machine learning models based on new data points. The server then adjusts the general model based on noisy gradients sent from the client devices.

Career Highlights

Pallika Kanani is currently employed at Oracle International Corporation, where he continues to develop innovative solutions in the realm of machine learning. His work focuses on enhancing privacy and security in collaborative learning environments.

Collaborations

Some of his notable coworkers include Michael Louis Wick and Adam Pocock, who contribute to the collaborative efforts at Oracle.

Conclusion

Pallika Haridas Kanani's contributions to machine learning and privacy preservation are noteworthy. His innovative patents reflect a commitment to advancing technology while ensuring data security. His work continues to influence the field and inspire future innovations.