Check out this Jupyter Notebook for my explorations of Reinforcement Learning applied to the game Easy21.

• One-Shot Learning Using a Deep Siamese Network

  An Experiment with One-Shot Learning

  As part of a job interview I had a few days to do a project to adapt a deep siamese network as described in the paper by Koch, Zemel an Salakhutdinov, and apply it to do one shot image recognition on the Olivetti faces dataset. I was pretty happy with the result so I am sharing it here.

• Learning Reinforcement Learning with Easy21

  Easy21 - A Reinforcement Learning Exercise

  Solutions to a Simple but Non-Trivial Problem

  The video recordings of David Silver’s Reinforcement Learning course are a great place to start if you want to understand Reinforcement Learning (RL), especially if you already have some coding experience. There is one assignment given out about half way through the course to which all RL approaches are applied. This is my write up of solutions to the assignment.