Seminar: Using Nearest Neighbor-Like Machine Learning Algorithms for Prediction of Transcription Fa
Ryan Zier-Vogel
Ph.D. Oral Comprehensive
Supervisory Committee: Dr. Lourdes Peña-Castillo, Dr. Todd Wareham and Dr. Andrei Igamberdiev
Using Nearest Neighbor-Like Machine Learning Algorithms for Prediction of Transcription Factor Binding Affinity
Department of Computer Science
Friday, Dec. 12, 2014, 11:00a.m., Room EN 2022
Abstract
I will use k-Neighborhood Components Analysis, Stochastic Neighbor Compression and Support Vector Machine-RankAggregation to predict the sequence preferences of one transcription factor based on the PBM data of related transcription factors. This task is important because gathering new transcription factor information is expensive and time-consuming so the ability to use previously existing information would save researchers time and money. A successful prediction algorithm is also needed because only ~1% of transcription factor binding preferences are known for eukaryotes