Projects
Private AI
In the Private AI project we try to solve the problem of data access problem. Due to privacy concerns, accessing sensitive data such as medical records and mental health related information is restricted. To solve this problem we develop ML algorithms that respect privacy by combining machine learning and deep learning algorithms with modern cryptography.
Behavioral Change Support
Our behaviors have large influence on our quality of life. Improper diets, smoking, hypertension, lack of physical exercise and other behaviors are contributing factors to many diseases. Furthermore, most treatments have behavioral components ranging from changes to diet and quitting smoking to taking medication on time. Our goal is to create technologies to support behavioral changes. Such technologies take into account differences between people and the setting they are at. In other words, we take personality into account when doing personalized medicine.
Fundamental ML algorithms
Our goal here is to develop algorithms to solve reoccurring problems in analyzing health related data. Such problems include explainability, finding sub-types of diseases, and more.