BIDS Senior Fellow and Chancellor’s Professor of Statistics Bin Yu and co-PI James Bentley Brown of LBNL have received a three-year NSF funding grant to develop statistical machine learning algorithms and associated uncertainty methods, along with scalable, open-source software and visualization tools. Their mandate is to enable the exploration of high-dimensional biological systems of previously intractable complexity in order to provide a framework for richly interpretable predictive models designed to help users understand emergent properties and dynamics.
All living systems - from cells to organs to the human body - are dynamic, interconnected and interdependent. They are composed of constantly interacting parts that exchange mass and energy with their environments at varying scales, and these interactions give rise to emergent phenomena that are difficult to predict from the dynamics of individual parts. The proposed work will take significant steps towards improving our capacity to learn about complex biological relationships among individual components and their subsequent interactions and behaviors. Proven principles and methods developed through this project will then be applied to other genomics problems in large scale computing experiments through the Amazon Web Service (AWS) platform.
More information can be found at this NSF website:
NSF Award #1741340 - BIGDATA: F: Scalable and Interpretable Machine Learning: Bridging Mechanistic and Data-Driven Modeling in the Biological Sciences