Research

My research interests mainly concentrates on two aspects: (i) eXplainable Artificial Intelligence (XAI) and (ii) optimizing machine learning models based on XAI for computer vision applications.

Machine Learning Theory: statistical machine learning, deep learning theory, Deep Neural Networks (DNNs), information theory, probability theory, and Bayesian theory.
Machine Learning Applications: neural network security, adversarial attack/defense techniques, image classification, image reconstruction, and image recognition.

I want to answer three questions of deep learning:

What is the functionality of DNN architecture, e.g., neuron, a hidden layer, and the entire DNN
What kind of information DNNs learn from dataset?
How to make DNNs better, especially improving the generalization and robustness performance?

Publications

Google scholar