Assistant Professor Basak Guler receives a National Science Foundation (NSF) CAREER Award to develop private, secure and trustworthy machine learning applications. NSF CAREER awards are awarded to assistant professors to fund research that is expected to form a firm foundation for a lifetime of leadership in integrating research and education. Collaborative machine learning allows multiple data owners to jointly train machine learning models, to improve AI-based systems by increasing the volume and diversity of data. While collaborative machine learning opens exciting possibilities, it can get tricky when used for privacy-sensitive data in the real world: things like healthcare records, financial transactions, or location data.
With this NSF grant, Guler aims to solve this problem by developing novel machine learning techniques that simultaneously protect the privacy of sensitive data. Typically, this has been limited by several major barriers, including the communication bottleneck – when there is not enough capacity to accommodate the current volume of traffic – security, and trustworthiness. Guler’s goal is to enable privacy-aware machine learning applications that feature three main characteristics: 1) accessible by users with limited bandwidth and computing power; 2) secure against adversaries; and 3) fair in its decisions towards all communities in society.