Hierarchical Models in Segway

by Adam Shaw for National Resource for Network Biology (NRNB)

Segway aims to find patterns in genomic data by segmenting the data, and assigning a small number of labels to each segment, and then using an unsupervised learning algorithm to discover patterns in the data. I aim to improve Segway’s ability to describe genome patterns by implementing features of a Dynamic Bayesian Network, specifically its ability to further segment into hierarchical labels, to reveal patterns at different scales.