NeurIPS 2021 Workshop Bridges Machine Learning and Physical Sciences
The NeurIPS 2021 Machine Learning and the Physical Sciences workshop unites computer scientists and physical scientists to tackle complex data-intensive challenges. The event focuses on applying AI to real-world scientific problems and using physical insights to improve machine learning techniques.
The NeurIPS 2021 Machine Learning and the Physical Sciences workshop brings together computer scientists, mathematicians, and physical scientists to explore how AI advances scientific discovery. Machine learning methods show great success in learning complex data representations, which enables novel modeling approaches across a wide variety of scientific disciplines at all scales in the universe.
Researchers at this workshop tackle data-intensive physical science tasks ranging from finding exoplanets in vast sky pixels to solving the quantum many-body problem and predicting extreme weather events. Critical techniques highlighted include segmentation, computer vision, sequence modeling, causal reasoning, generative modeling, and probabilistic inference.
Beyond simply applying machine learning models to scientific challenges, the workshop places a strong emphasis on interpreting what these models actually learn. By fostering this interdisciplinary dialogue, the organizers expect to introduce exciting new open problems to the broader community and stimulate the creation of innovative approaches for solving difficult scientific questions.