Reinforcement learning is one of the most popular approaches for automated game playing. This method allows an agent to estimate the expected utility of its state in order to make optimal actions in an unknown environment. We seek to apply reinforcement learning algorithms to the game Flappy Bird. We implement SARSA and Q-Learning with some modifications such as $\epsilon$-greedy policy, discretization and backward updates. We find that SARSA and Q-Learning outperform the baseline, regularly achieving scores of 1400+, with the highest in-game score of 2069.

At the height of its popularity, Flappy Bird was possibly the biggest waste of time humanity indulged in. Luckily the age of artificial intelligence is coming and we can offload the mundane tasks to artificial intelligence. Let's train an AI to play Flappy Bird, so we don't have to. Note: This article is for people who already have a basic idea about working with Unity and are interested in Neural Networks and Reinforcement Learning. No prior experience with neural networks and/or PhD required!

First impressions are notoriously subjective (and flawed), but now machines are being trained to make similar snap judgments based on human-generated data. AI-published books may not be too far away! Researchers at Kyushu University in Japan have trained a deep neural network to study book covers and determine their category. On a hunt for interesting and high-quality datasets to use for machine learning, I stumbled upon these 20 weird and wonderful sets. Neat! Check out this open source deep convolutional neural network that is trained on transformed audio signals to recognize "ahem" sounds.