| |
| Computing and Software Systems University of Washington, Bothell |
 |
J. Jeffry Howbert, Ph.D. Affiliate Professor Computing
& Software Systems, University of Washington
Bothell |
| My research interests center on the application of machine
learning to biomedical problems.
In the past few years I have pursued these interests at
several companies and an academic lab. At NeuroVista Corp., I worked on the algorithms team, developing models for real-time prediction of seizure risk from the EEG of epileptic patients. At Insilicos LLC, I developed a platform for parallel processing of machine learning ensembles in the Amazon cloud. I also worked on discovery of biomarkers for cardiovascular disease, using data from mass spec proteomic experiments. In Bill Noble's lab at University of Washington, I am using machine learning to develop new scoring algorithms for peptide-spectrum matches in shotgun mass spec proteomic experiments. I also consult for Zillow.com in the quantitative modeling group, working on improvements to their core real estate valuation algorithms. A few years ago, I was a leading competitor in the Netflix Prize contest. Our team, The Ensemble, finished second in a very close race. For most of my professional life I did research in a different field, drug discovery. You can see a synopsis of those efforts under my publications and at LinkedIn. |