Curriculum Vitae

Link to CV.

Research Interests

My research focuses on non-convex optimization for spectral methods (matrix and tensor decomposition) and learning latent variable graphical models (such as latent Dirichlet allocation, mixtures of graphical models, hidden Markov models) on distributed systems with large-scale data. Some of my previous works involve distributed spectral decomposition techniques for topic modeling and mixed membership detection for large-scale networks with extension to temporally evolving networks. 

I am also interested in optimization of large scale numerical algebraic operations and distributed computing systems. For instance, I worked on distributed realization of alternating minimization for tensor decomposition on the cloud.

Another thread of my research lies in computational biology and neuroscience. Generally I am interested in applying machine learning techniques to help target biological experiments. 

Some examples of the research I do:

- efficiently locating the optimal solution for high-dimensional non-convex functions such as finding tensor decomposition with global convergence guarantee;

- developing fast detection algorithm to discover hidden and overlapping user communities in social networks;

- designing a parallel spectral tensor decomposition algorithm on Map-Reduce frameworks for automatic categorization of articles;

- learning convolutional sparse coding models for understanding semantic meanings of sentences and object recognition in images;

- healthcare analytics by learning a hierarchy on human diseases for guiding doctors to identify potential diseases for patients;

- inferring brain cell types and gene expression profiles under different cell types to understand the brain.