Add initial EBImage Machine Learning Baseline

[gwaybio]

Using the EBImage features (processed in 2.process-data), what is the baseline performance in the testing set?

Add documentation about how to reproduce the analysis. I believe that @wagenrace is working on a random forest implementation 🎉

[gwaybio]

It will be important to output several performance metrics, so that we can develop visualization strategies as described in #20