Congratulations to computer science graduate student James Hughes who was one of four winners of the Graduate Poster Session held recently at the Top of the Hop! Enjoy your winnings, James! (Read on for a summary of Hughes’s poster.)
“Nonparametric Sparsification of Complex Multiscale Networks”
Scientists deal with massive amounts of data. Analyzing data often involves comparing two objects of study by quantifying some relationship between them (e.g., friendships in social networks or correlations between stock prices over time). Many of these interactions are not significant and can be detrimental to subsequent scientific analysis. We propose a method to remove the insignificant interactions in the data by applying ideas from the analysis of networks, in which we consider local interactions and retain only those that are highly probable given the data.
poster summary by James Hughes
photo by Erin O’Flaherty