Machine learning(1): Supervised learning, Unsupervised learning and Reinforcement learning

Machine learning is divided into three types of method learning.
The first method is supervised learning and the second is unsupervised learning. Finally, it is reinforcement learning.

Supervised Learning

Characteristics of supervised learning:

1. Each tuple in the training set is labelled or targeted.
2. The test samples are also labelled or targeted.
3. Use methods of regression to fit the actual labelled data.

Unsupervised learning

Characteristics of unsupervised learning:

1. All the training set is not labelled.
2. Use methods of clustering:
   <1>points within each cluster are similar to each other.
   <2>points from different clusters are different.
   This requires the definition of a meaningful similarity (or distance) measure.
3. Association rules:
   Association rules are simply statements that certain groups of items or events tend to occur together.

Reinforcement learning

Characteristics of reinforcement learning:

1. We get feedback/reward only after a complete sequence of actions/decisions.
2. There is no explicit (not even retrospective) scoring of individual actions/decisions
3. The goodness of each action depends on the actions that follow (sometimes on the opponent’s/environment’s response to our action).