Research conducted by Tongliang Liu. If you are interested in the research and would like to have a discuss with me, please feel free to contact me.
Statistical learning theory
"There is nothing more practical than a good theory."
- Kurt Lewin
Project: hypothesis complexity measurement
Statistical learning theory provides the mathematical and theoretical foundations for statistical learning algorithms and inspires the development of more efficient methods. We are interested in how to measure the complexity of the algorithmic hypothesis class, which is a subset of the predefined hypothesis class that a learning algorithm will (or is likely to) output. Finding such a small subset will give us insights on explaining how learning algorithms work.
Learning with label noise
"Labelling label is becoming more and more difficult as the data size growing bigger exponentially."
Project: noise rate estimation
We assume that the true labels of examples flip into the corrupted labels with some probabilities, with which the true labels can be inferred from the noisy ones.