Tag: ai

AlphaGo: what it does and why it works

In 2016, the team led by David Silver and Aja Huang released Alpha Go, the first computer program that beat a professional human player at Go. It also beat every engine ever made before it by a large margin. Their approach is detailed in the paper Mastering the game...

RL algorithms – the key ideas summarized

We assume you know the basic ideas of reinforcement learning: agents, actions, rewards, environment, episodes, total episode reward, discounted reward, Q-values, state values. Over the history of reinforcement learning, tens of algorithms were developed and it can be difficult to keep track of all the progress. As a refresher,...