Research Direction: My general research interests are network science, machine learning, data mining, computational complexity theory, and their applications in biology.

My current research interests include machine learning for classification and dimension reduction, and their computational complexity analysis. The focus of this line of research can be divided into two aspects. The first aspect is designing linear classifiers applicable to the classification tasks where the data are significantly affected by outliers. The second aspect is designing low-complexity linear classifiers that is able to achieve comparable classification performance to the commonly-used machine learning methods. Moreover, a few classification methods can be considered to comprise different orders, and can be approached by only part of these orders. Generally, the complexity will decrease at the cost of omitting some non-substantial orders that gains little on the performance. As the characteristics of a classifier will change after being approximated, it worth further investigations what pros and cons it will bring if we only keep the orders that make main contributions.