Detailed Balanced Chemical Reaction Networks as Generalized Boltzmann Machines
Tune in for the live stream on YouTube or Twitter
Abstract: Can a micron sized sack of interacting molecules understand, learn, and adapt to a constantly-fluctuating environment? Cellular life provides an existence proof in the affirmative, but the principles that allow for life's existence are far from being proven. In this talk, I will draw insights from machine learning theory, chemical reaction network theory, and statistical physics to understand how various classes of chemical systems can implement and be interpreted as probabilistic inference machines. These results illustrate how a biochemical computer can use intrinsic chemical noise to perform complex computations and provide examples of well-specified physical systems capable of representing complex distributions and autonomously learning from the environment.