Read the HOPE model paper.
Collect the character statistics of a news corpus to do classification. Libsvm is still running.
Modify iKids code.
Start reading variational inference.
Wasted a lot of time.
It seems that I will not publish paper here, I may quite soon.
Budha belss me.
原文：http://www.cnblogs.com/peng-ge/p/4322565.html