I have executed several interesting machine learning projects, including training models using state of the art methods in Deep Learning.
Spot Nuclei.
Speed Cures. 
For instance, my group came in 221 out of 816 teams in the 2018 Kaggle Data Science Bowl on the second stage of the competition using our Mask-RCNN Deep Learning Model.
The goal of the competition was to train a model that can automatically detect cell nuclei from slides, which doctors must currently do manually, thus allowing for much accelerated research in cancer and rare diseases.
All information (including my other team members) are included in the below paper and slideshow.