Heli Ben Hamu is pursuing her PhD in the Department of Computer Science and Applied Mathematics under Prof. Yaron Lipman’s supervision. Her research focuses on geometric deep learning, an exciting field with significance for artificial intelligence (AI).
Deep learning utilizes neural networks, algorithms that mimic the most simplified model of the human brain, exceeding human performance of many tasks, including the interpretation of 2D visual images. On the other hand, geometric deep learning focuses on more complex data types, such as graphs and 3D models. Heli’s research focuses on designing and gaining a better theoretical understanding of deep learning algorithms applied to graph data. Graphs are a flexible representation for many kinds of data such as molecules, social networks, scene understanding, etc. Creating algorithms that improve computerized systems’ ability to infer and extract information from such data types has a wide span of applications to real-life problems such as drug design, navigation systems and fake profiles detection on social networks.
In her free time, following her daughter’s birth, Heli has initiated the first “Coping with the challenges of early motherhood and graduate studies” workshop in cooperation with the Weizmann Institute social services and she serves as a mentor in the Mom2mom program.