Mercye assignment 9

[mercye]

9_1

It was hard to create the cv function because I struggled to split the data set into features and targets after zipping the two together in order to shuffle, and then I struggled to split the data into test and training. I understand the cross_validation function more in that I have clearer questions about what is scoring. More specifically, why does the .score take the test data and true target labels as arguments rather than the test data and the predicted label arguments? In any case, I think that I have a good base understanding how cross-validation, how to interpret the results, and what its purpose is.

9_2

I truly do not think I successfully built a random tree classifier. I understand that random trees are a collection of decision trees averaged together in order to build a better predicting model; however, I did not understanding how to loop so that the predictions from decision trees are able to be "averaged" together.

9_3
The text classifier is super cool. I don't have any projects, currently, that I would like to do with text classification; however, I could imagine myself using this with tweets from the twitter streaming api. I also really like walking through code like this.