MIL - The University of Tennessee
the university of tennessee machine intelligence lab
Tomer Lancewicki, Postdoc
[ resume
Machine Intelligence Lab
606 Min H. Kao Building
The University of Tennessee
Knoxville, TN 37996-2100
Research Deep Learning Architectures 
  My research focuses on efficient ways of improving the inference process in deep learning architectures, particularly through approximation of the covariance matrix in non-stationary settings.
Selected Publications
[1]     B. Goodrich, T. Lancewicki, I. Arel, "Sequential Covariance-Matrix Estimation with Application to Mitigating Catastrophic Forgetting," to appear in IEEE International Conference on Machine Learning and Applications (ICMLA), December, 2015 [pdf]
[2]     T. Lancewicki, I. Arel, "Regularization of the Kernel Matrix via Covariance Matrix Shrinkage Estimation," to appear in Proc. the SIAM Conferences on Applied Linear Algebra, October, 2015
[3]     T. Lancewicki, I. Arel, "Covariance Matrix Estimation for Reinforcement Learning," in Proc. of the 2nd Multidisciplinary Conference on Reinforcement Learning and Decision Making , June, 2015 [pdf]
[4]     T. Lancewicki, "Multi-target Shrinkage Estimation for Covariance Matrices," in Proc. of the 39th SIAM Southeastern Atlantic Section Conference, March, 2015