What is Reinforcement Learning in Machine Learning?

Reinforcement Learning

Reinforcement learning is the training of Machine Learning models to make a sequence of decisions.

• It is about taking suitable action to maximize reward in a particular situation.
• It is based on the Reward and Policy principle.
• Agents perform actions, in response gets reward/penalty, the state will be changed, and based on that policy will be made.
• This type of algorithm generally used in game development.

Examples

• Baby learning how to walk: In this example, the child is an agent, whereas the environment is the surface on which they walk. Here, the baby is taking action and is moving from one place to another. In other words, it is changing its state. If the child is able to walk, then he gets a rewardtoy; otherwise, he will not get any reward.
• A dog receives a reward from a trainer for completing a certain task and no reward for failing to do so.
• Making a humanoid robot walk:
  • positive reward for forwarding motion
  • negative reward for falling over

Framework of Reinforcement Learning

In a way, Reinforcement Learning is the science of making optimal decisions using experiences. Breaking it down, the process of Reinforcement Learning involves the following steps:

• Observe the environment
• Figure out how to act using some strategies
• Act accordingly
• Receive a reward or penalty
• Learn from the experience and refine our strategies
• Iterate until an optical strategy is found