Machine Learning With Molecular Biology Protocols

 

Prof. Russell Deaton

University of Arkansas

 

October 1, 2010 1:00 PM

302-209

 

The promise of DNA computing is the vast amount of information that can be stored in microscopic volumes. The challenge is to overcome the numerous errors and randomness that are inherent to in vitro processing of DNA. One way around this challenge is to look for algorithms that are tolerant or take advantage of randomness. Many such algorithms are part of Machine Learning. In this talk, several molecular biology protocols that implement Machine Learning algorithms will be summarized. Specifically, a learning protocol, which was inspired by associative memories, that learns an unknown input DNA sequence, and a selection protocol, which was inspired by genetic algorithms, that selects noncrosshybridizing sequences will be presented. Finally, potential for future applications to Hypernetworks will be discussed.

 

This page is maintained by Ji-seon Yoo (jsyoo@bi.snu.ac.kr).
Last update: September 9, 2010