reading co-trainning papers

I think Minmin's co-training paper did  a decent job. It's clearly presented. And from this, I can see that if we want to be a good researcher, something we have done is:

1. Read plenty of papers, especially the ones related to your research area. 

2. Understand different problems and have a clear picture what problems have what kind of state of art.

3. Learning from the papers, and try to do and implement it. 

4. Question what the other people do and question about your own methods.