Supervised & UnSupervised & Reinforcement Learning

The Output of the Model is

• Label

• Feature

When the Model Have The Input and Output this phase is called

• Training

• Testing

• Prediction

When the model Only have the Input

I Don't know the real Output

• Training

• Testing

• Prediction

minimize the error as possible

• Prediction

• Optimization

• Training

• Testing

Low accuracy of training

Low accuracy of testing

• Fitting

• Over fitting

• Under Fitting

High accuracy of training

Low accuracy of testing

• Fitting

• Over Fitting

• Under fitting

this model split the Data to 3 Groups

• Classification

• Clustering

the Code in Supervised & Unsupervised Called

• Program

• Model

• Agent

the Code in Reinforcement Learning Called

• Model

• Agent

• Program

What is "Environment"?

• It is the place that the Agent take actions in and it responds with positive or negative rewards

• It is the Reward itself that the Agent learns from