{"id":7323,"date":"2016-04-27T05:44:00","date_gmt":"2016-04-27T05:44:00","guid":{"rendered":"http:\/\/www.kurzweilai.net\/?p=278886"},"modified":"2016-04-29T05:44:39","modified_gmt":"2016-04-29T05:44:39","slug":"public-beta-of-toolkit-for-developing-machine-learning-for-robots-and-games-released","status":"publish","type":"post","link":"https:\/\/hoo.central12.com\/fugic\/2016\/04\/27\/public-beta-of-toolkit-for-developing-machine-learning-for-robots-and-games-released\/","title":{"rendered":"Public beta of toolkit for developing machine learning for robots and games released"},"content":{"rendered":"
\"\"<\/a>

Make a three-dimensional bipedal robot walk forward as fast as possible, without falling over (credit: OpenAI Gym)<\/p><\/div>\n

OpenAI (a non-profit AI research company sponsored by Elon Musk and others) has released the public beta of\u00a0OpenAI Gym<\/a>, a toolkit for developing and comparing algorithms for reinforcement learning\u00a0(RL), a type of machine learning.<\/p>\n

OpenAI Gym consists of a growing suite of\u00a0environments<\/a>\u00a0(from simulated robots<\/a>\u00a0to\u00a0Atari<\/a>\u00a0games), and a site for\u00a0comparing and reproducing<\/a>\u00a0results. OpenAI Gym is compatible with algorithms written in any framework, such as\u00a0Tensorflow <\/a>and\u00a0Theano<\/a>. The environments are initially written in Python (other languages planned).<\/p>\n

If you’d like to dive in right away, you can work through a tutorial<\/a>, and you help out while learning by\u00a0reproducing\u00a0a\u00a0result<\/a>.<\/p>\n

What is reinforcement learning?<\/strong><\/p>\n

Reinforcement learning (RL) is the subfield of machine learning concerned with decision making and motor control. It studies how an agent can learn how to achieve goals in a complex, uncertain environment. It’s exciting for two reasons, according to OpenAI’s Greg Brockman and John Schulman:<\/p>\n