Vector Lecture Series - Kyunghyun Cho скачать в хорошем качестве

Vector Lecture Series - Kyunghyun Cho

Talk title: Causality Meets Deep Learning Speaker: Kyunghyun Cho | Professor of Computer Science & Data Science, New York University; Research Scientist at Genentech Abstract: In this talk, Kyunghyun Cho explores the relationship between causality and association, emphasizing why these concepts should be understood as complementary rather than competing. He breaks down what we mean by causality, causal discovery, and causal inference, highlighting the distinction between uncovering underlying mechanisms and discovering causal relationships. Cho then examines the challenges of performing causal inference and causal discovery in practice—why we often struggle, and where traditional estimation methods fall short. The talk concludes with a look at emerging deep learning–based approaches to causal inference and discovery, which significantly reduce the need for developing new estimation algorithms and may reshape how researchers tackle causal questions in complex systems. About Vector’s Distinguished Lecture Series: The Vector Distinguished Lecture Series brings together leading academic and industry researchers across the Greater Toronto Area (GTA) to explore cutting-edge topics in machine learning, artificial intelligence, and data science. The goal of the series is to foster collaboration, spark new ideas, and strengthen the machine learning community in Toronto and beyond. All talks are open to the public and streamed online for remote attendees.