DML Weekly Report

1. What you have done
* Helped Yao with his rough draft, get into our code base, etc.
* Discovered how to use weka classes in my code, serialize models, and then reload the models and run "production" tests/questions on it.
* Found 5 papers related to our 3 papers
* Wrote a lot of code

2. What problems you have encountered
* I am struggling to understand how to get probability scores out of a classifier. For example, if I train a boolean classifier, I want it to output a score between 0 and 1 instead of just o or 1 so I can do my own threasholding.

* I (as always) need a day or two more to get you a rough draft of my paper, and then another week or two to start getting results.

3. What possible solutions you are considering for these problems
* I'll ask you about the first and the second.... well just throw myself upon your good graces yet again....

4. What you plan on doing in the coming week
* Work more on media items for HS presentation (video)
* Finish my abstract, help Graham finish his, get probability scores out of weka, finish fast evaluation system.


5. New ideas, specific topics/issues you wish or us to focus on in
I need to be able to get out of a NN, or Decision tree a probability score (even on new instances) rather than a Y/N answer/class.