 |
Center for Nonlinear Studies |
I will talk about the “interactive and transparent machine learning” research we do at the Machine Learning Laboratory (http://ml.cs.iit.edu) at Illinois Institute of Technology. Specifically, I will talk about i) how transparency lead to the discovery of a novel and empirically quite effective active learning algorithm, ii) simple, intuitive, and effective interactive learning algorithms that drastically improves model performance, and iii) active inference algorithms that can save resources during the use of the model for decision making. While theoretical aspects will be mentioned as needed, the primary focus of the talk will be algorithms and empirical results.
Host: Diane Oyen