Learning & Modeling Diversity
Learn about Determinantal Point Processes and how they can be used to model diversity in neural networks, texts, and recommender systems. Even as we become able to access larger and larger amounts of data, we typically wish to access it in an diverse fashion. Although this ability is useful for complicated tasks like automatic text summarization, we also require it on a day-to-day basis, such as when using search engines (e.g., when looking up "python," this allows us to find information on programing languages and snakes).
Time
10:00 AM to 3:00 PM
Sponsor
Laboratory for Information and Decision Systems