We have developed novel approaches to tackle high dimensional and dynamical data. We have involved in many topics such as SVM, ICA etc. In recent years, we are in particular interested in exploring structures such as communities in a give data set . Another issue is about Lp regularization, where 0<p<1. We pioneeringly introduced Young measure to deal with the issue and proved that the solution is localized and has many nice properties.
Figure: Community structures revealed using ‘local mixing’ property of a Markov chain. For details, see the following paper 
 Yang B., Liu JM, Feng J.F.,Liu DY(2010) On the spectral characterization and scalable mining of network communities IEEE T. Knowledge and Data Engineering