MIL - The University of Tennessee
the university of tennessee machine intelligence lab
  
Aaron Mishtal, Ph.D. Student
  
Machine Intelligence Lab
606 Min H. Kao Building
The University of Tennessee
Knoxville, TN 37996-2100
  
Research Deep Learning Architectures 
  
  Neural Networks
  My current work focuses on optimization techniques for concurrently trained and operated ensembles of neural networks. In particular, I study how information-theoretic metrics can be used to enrich diversification in such settings.
  
Selected Publications
  
[1]     S. Young, I. Arel, A. Davis, A. Mishtal, "Hierarchical Spatiotemporal Feature Extraction using Recurrent Online Clustering," Pattern Recognition Letters, February, 2014 [pdf]
[2]     R. Coop, A. Mishtal, I. Arel, "Ensemble Learning in Fixed Expansion Layer Networks for Mitigating Catastrophic Forgetting," IEEE Trans. on Neural Networks and Learning Systems, pp. 1623 - 1634, October, 2013 [pdf]
[3]     A. Mishtal, I. Arel, "Jensen-Shannon Divergence in Ensembles of Concurrently-Trained Neural Networks," in Proc. IEEE International Conference on Machine Learning and Applications (ICML), December, 2012 [pdf]