Abstract: 

Complex AI applications, such as disaster preparation and mitigation or predicting the effects of climate change on species survival, require multiple paradigms, including data-driven predictive models, physics simulations, and probabilistic models. Hybrid AI is an emerging field that integrates model components from different paradigms in a unified model for complex applications like these. However, existing hybrid AI frameworks are usually ad-hoc, specific to certain configurations of models, and lack explainability, which is vital for real-world applications. We have developed a new hybrid AI framework called Scruff, based on probabilistic programming principles, that provides a coherent, general, and explainable way to build multi-paradigm and multiscale models. Scruff is implemented in Julia and available open source on GitHub.

In this workshop, I will explain the core principles of Scruff and the main programming concepts. I will then demonstrate how we used Scruff to create a tool for wildfire risk assessment and mitigation that includes climate models, historical fire data, and fire propagation simulators. Finally, we will work through a hands-on session of getting up and running with Scruff and implementing and running simple models.

Session Outline:

Module 1: introducing hybrid AI and Scruff and its main concepts
Module 2: demonstrating Scruff through a wildfire risk assessment application
Module 3: downloading, installing, and getting up and running with Scruff (Julia installation is a prerequisite)
Module 4: building and running simple models

Background Knowledge:

Julia; probabilistic models such as Bayes nets; familiarity with probabilistic programming helps, but no specific language is required

Bio: 

Dr. Avi Pfeffer is Chief Scientist at Charles River Analytics. Dr. Pfeffer is a leading researcher on a variety of computational intelligence techniques including probabilistic reasoning, machine learning, and computational game theory. Dr. Pfeffer has developed numerous innovative probabilistic representation and reasoning frameworks, such as probabilistic programming, which enables the development of probabilistic models using the full power of programming languages, and statistical relational learning, which provides the ability to combine probabilistic and relational reasoning. He is the lead developer of Charles River Analytics’ Figaro™ probabilistic programming language. As an Associate Professor at Harvard, he developed IBAL, the first general-purpose probabilistic programming language. While at Harvard, he also produced systems for representing, reasoning about, and learning the beliefs, preferences, and decision making strategies of people in strategic situations. Prior to joining Harvard, he invented object-oriented Bayesian networks and probabilistic relational models, which form the foundation of the field of statistical relational learning. Dr. Pfeffer serves as Action Editor of the Journal of Machine Learning Research and served as Associate Editor of Artificial Intelligence Journal and as Program Chair of the Conference on Uncertainty in Artificial Intelligence. He has published many journal and conference articles and is the author of a text on probabilistic programming. Dr. Pfeffer received his Ph.D. in computer science from Stanford University and his B.A. in computer science from the University of California, Berkeley.