Here is a link to my Github.
I'm pretty good at coding in Python. Some libraries I've used before are: TensorFlow, pandas, numpy, scikit-learn, Gurobi, cvxpy, NetworkX, and scipy. I also know my way around SQL: Postgres and PostGIS (for geo-spatial data).
During the summer of 2018 I interned at Roadie as a Data Scientist where I helped develop algorithms to notify drivers of packages they might be interested in. I helped with the design, implementation, and testing of the algorithms. By the end of the summer I had 5500+ lines of code in production.
The figure below shows a comparison of Saddle Point Follow the Leader vs other algorithms for training GANs in a mixture of Gaussians dataset.
Below you can see some videos on how I taught Hopper how to move using a Reinforcement Learning algorithm called DAGGER. You can find the code and more details on my Github.
The video in the left shows Hopper after 8 training epochs, to the right you can see him after 16 training epochs.