Project:

In this project, I will use slow intelligence system to model feature
selection problem solving and to design time controller for this problem.
One state-of-the-art literature [1] proposed a new method to solve such
problem, which is proven to have good statistical property although not
very efficient. I will model this method to five phases of SIS and
introduce our new idea about Knowledge base and time controller of SIS.
There are two tasks in the project.  Firstly, I will use SIS to model the
existing method in mathematical way. Second, I will design time controller
in term of Petri Net and introduce Knowledge base in this system.

In next step, beyond the project, I will use some real data and synthetic
data to do the experiment and compare the results with some existing
feature selection method like LASSO, forward, backward regression, etc.
Moreover, I will use some visualization tool to visualize the result and
the process of feature selection.

Reference:
[1] “Ultrahigh dimensional feature selection: beyond the linear model”,
Jianqing Fan, Richard Samworth and Yichao Wu
[2] “ A General Framework for slow intelligence systems”, Shi-Kuo Chang
[3]” Modeling Human Intelligence as A Slow Intelligence System” , Tiansi Dong


Comments: If possible, use pnml to model the SIS application
so that you, Xu and Shaw can shaw the same specification
language.