Matthew Wallace

Matthew Wallace is a Computer Science and Applied Mathematics double major at Texas State University.  He is interested in data-mining, machine learning and artificial intelligence.  He is currently working under the guidance of Dr. Byron Gao.

Current projects:

  • A novel similarity measure allowing for mitigation of the curse of dimensionality when working with high dimensional data sets.  Not only does such a measure not exclude very many of the current tactics for handling high dimensional data but these existing ideas can be incorporated into the whole process.  Since it is a similarity measure and not an actual algorithm it can be used with numerous existing algorithms.  Current experimentation with this measure is being conducted using KNN.
  • A lazy classification algorithm that allows near training examples to block the "visibility" of training examples that are further away.  Influence of the so called visible training examples is also weighted on distance allowing closer examples to more strongly influence the outcome.  The initial goal is to better identify clusters that are embedded inside other clusters that could otherwise be mislabeled.