One of the critical shortcomings of big data-driven deep learning is its black-box nature. To help resolve this, it’s important to develop architectures and algorithms that can capture the fundamentals of how humans learn and infer. Join Professor Hanwang Zhang from Nanyang Technological University in Singapore, Microsoft Senior Researcher Yuwang Wang, and Professor Shujian Yu from the Arctic University of Norway as they share their work and insights on how to achieve interpretable learning by leveraging representation disentanglement and information theory. You’ll learn about these powerful concepts and discover how they help address interpretability and generalization in deep learning.
Learn more about the 2021 Microsoft Research Summit: https://Aka.ms/researchsummit