YouTube Video Review: Learn Data Science by Doing Kaggle Competitions: Connect X

There are three types of machine learning, being supervised, unsupervised, and reinforcement learning. I have written a lot about supervised learning because most of the datasets I have worked with have had labels for me to work with. I have also written a bit about unsupervised learning, and even a few times about semi-supervised learning.

One subject that I have not really written about is reinforcement learning, and the reason for that is because there is so much more material available about supervised learning than reinforcement learning.

Reinforcement learning is a machine learning training method based on rewarding desired behaviours and/or punishing undesired ones. In general, a reinforcement learning agent is able to perceive and interpret its environment, take actions and learn through trial and error.

In reinforcement learning, programmers devise a method of rewarding desired behaviours and punishing negative behaviours. This method assigns positive values to the desired actions to encourage the agent and negative values to undesired behaviours. This programs the agent to seek long-term and maximum overall reward to achieve an optimal solution.