Reinforcement learning is an area of machine learning that focuses on how to design a software program that will grow into a system. This is done by guiding it with punishments and rewards.
Analogy:
- To get a tree to grow into a certain shape we physically guide the sapling in its lifetime development.
- A dog is trained with the correct application of positive reward and negative punishment in its training.
The environment the software finds itself is expressed as a Markov decision process.
Author: Seamus Taylor
Post #: 6
No comments:
Post a Comment