13 Mar 2016 - Experiment Suggestions

It’s been a long time that I have some spare time to squander. However, I actually have no idea about what to do when not being busy. So I just summarize the past several weeks to practice my writing.

Two weeks ago, I was doing the assignment of Linear Models. There are still three problems remained unsolved in that Sunday evening. Thanks to a friend in Department of Statistics for sending me the solution. After submitting the assignment of Linear Models on Monday, I turned to the assignment of Matrix. Luckily I found all the answer via Google (=_=). Last weekend, I spent two days leaning SNAP and made slides for the tutorials. And this week was spent mostly to prepare for the group meeting on Friday. Finally I can have a rest tonight.

The courses I enrolled this semester are a little bit hard, so there will a much tougher period at the end of this semester. Wish I could survive. What mostly pleased me was the research work I did this week since it clearly shows several possible directions to go next.

1. Causal cut on model without confounders

2. Try fully connected network

3. Add confounder in one group and check results

4. Modify the cut finding process

5. Tune parameters

Wish that I could insist on writing lousy things like this on Jianshu.