Home

Research

Research I've done either for class, my work or on my own, mostly in Machine Learning.

  • Masters Thesis
    Tuesday, August 24, 2010 - 07:54

    In mid July 2010 I defended my thesis. The defense was followed by a rush to get my thesis revised and submitted by July 31. After the mad dash, a committee member that had more revisions than all the other four combined, and barely getting that member to sign in time, I finally printed my thesis and had it submitted to the graduate college. And yes, you read that correctly, I had five committee members on my ms committee.

  • The Search for a Splitting Statistic
    Sunday, April 11, 2010 - 15:02

    My research involves decision trees and algorithms that grow them. In order to do this we need a splitting statistic that can tell the quality of a split. Our trees are binary where each node asks a "Yes/No" question. This means splits produce a contingency table like this (for 2-class data).

  • Calculating Minimum Volume Bounding Box
    Thursday, February 25, 2010 - 12:31

    In my work I ended up needing to know the minimum volume bounding box (MVBB) of a given of points in 3d. This is also called the oriented bounding box (OBB). Finding the axis-aligned bounding box (AABB) is a simple task, however it can have a volume much larger than needed. Come to find out, finding the minimum volume bounding box is a very difficult problem with several techniques for finding it. However, I didn't find any useful code for finding the MVBB.

  • Tight Convex Hulls: Who Are My Neighbors?
    Monday, February 2, 2009 - 17:00

    In one of my research projects we have a set of points on a two dimensional plane that when given a single point from that set we want to be able to know who its "neighbors" are. For our work we have the mesonet stations spread across the state of Okalhoma and we want to be able to select the neighboring stations. There are a couple of methods that are intuitive but require setting a parameter, something I'm not fond of. Also we have the concept of multiple levels of neighbors, so my first neighbors, my second neighbors and so on.

  • Brittleness in Spatio-Temporal Relational Probability Trees
    Monday, December 15, 2008 - 12:56

    The Fall semester of 2008 I started my masters at the University of Oklahoma. My first project which has dovetailed with a class and my work as a graduate research assitant has been a study of brittleness in spatio-temporal relational probability trees. Expect to see my class paper up here soon.

    So a year and half later I finally put up my class paper.

  • Generalized Grammatical Evolution
    Thursday, December 11, 2008 - 10:56

    I've got some cool ideas for generalizing grammatical evolution. Expect to see them here as the mature.