Types of Reinforcement Learning

Reinforcement Learning is broadly divided into two parts:

1. Positive reinforcement

2. Negative Reinforcement

1. Positive Reinforcement

1. This is adding something that strengthens the behavior of an agent.
2. It increases the tendency of a certain behavior to occur again.
3. Also, Positive Reinforcement is further divided into two parts:
   1. Primary Learning
   2. Secondary Learning

PRIMARY LEARNING

This type of reinforcement learning is called unconditional reinforcement which fulfills the basic need of the agent. The basic needs include food, water, air, sex, etc. This type of reinforcement is not learned by the agent.

SECONDARY LEARNING

This includes actions that strengthen a certain behavior in association with Primary Learning. For example, a child starts scoring well every time you give him his expensive game, and you praise him by saying 'good boy. Over the course of time, you give him less expensive games with the same praise, but still, he keeps on scoring good marks. A time will come when you give him nothing but just 'praise', and the child will still score well. So here, the expensive gift will act as Primary Learning and praise will be secondary learning.

2. Negative Reinforcement

1. This type of reinforcement is adding strength to behavior by avoiding or removing negative actions or stimuli.
2. Let's understand the concept with the help of an example. For example, an employee thinks that he needs to complete his presentation by the current night, or else he will be scolded by his boss the next day. So, now to avoid the scolding, he will work hard, to complete his presentation. The work that he will do to avoid the scolding is a part of Negative Reinforcement