Merge branch 'master' of github.com:microsoft/datascience4managers

Part_1/ML_classifier_results.md

@@ -0,0 +1,4 @@
+These plots show the results of training an ML model on the same training set used for selecting rules, and evaluating it on the same test set. The TF/IDF of the words from the posts taken one at a time (unigrams) or two at a time (bigrams) were used as generate features, which were run through a LogisticRegression classifier in the Python `slkearn` library.
+
+![](images/figure-markdown_strict/ML_confusion_matrix.png)
+![](images/figure-markdown_strict/ML_ROC_curves.png)