* Title: Noise Benefits in Markov Chains * Abstract: Markov chains underlie important models in physics, chemistry, economics, and data mining. I will discuss results that show for the first time that noise can benefit a very general class of Markov chains. I will show how noise can drive Markov chains to explore novel solutions and speed convergence to steady-state. During the talk I will demonstrate the noise benefits results over three Markov models. The first shows noise benefits in the Ehrenfest diffusion model and illustrates the noise-benefit mechanism in the large class of birth-death processes. The second uses a noisy Wright-Fisher stochastic population-genetics model to show a faster time to find the steady-state genomic probabilities. The third uses noise in an empirical Zeolite synthesis pathway to show that noise benefits may exist even in an empirically derived Markov process. The talk will cover fundamental Markov chain concepts and extend the results into future areas of application such as Markov chain Monte Carlo (MCMC), benefits for massive Markov models such as the Google PageRank algorithm, and new research avenues to study the profound possibility of speeding evolution in preferred directions. * Bio Brandon Franzke received a joint B.S. degree in Biomedical and Electrical engineering with a minor in neuroscience from the University of Southern California (USC) in 2002. He is finishing Ph.D. studies at USC on noise benefits in statistical signal processing under the supervision of Professor Bart Kosko. His efforts focus on profound applications of noise in fundamental arenas of chemistry, biology, and physics and his work bridges formal theory to the experimental laboratory.