Reinforcement Learning

Mario in Super Mario bros is considered

• Agent

• State

• Action

• Reward

........ Is used to evaluate Agents' Actions.

• Agent

• State

• Action

• Reward

We prefer to use ............ when you have Features and Labels

• Supervised Learning

• UnSupervised Learning

• Reinforcement Learning

We prefer to use ............ when you have Features Only and *No* Labels

• Supervised Learning

• UnSupervised Learning

• Reinforcement Learning

We prefer to use ............ when you have no features neither labels but you can build a reward equation.

• Supervised Learning

• UnSupervised Learning

• Reinforcement Learning

We need to use Gamma in our cumulative reward to .....

• Discount near rewards

• Discount far rewards

What should be done first .........

• Exploration

• Exploitation

Exploitation is

• Doing random actions to know more about the environment.

• Using Current knowledge to do best actions

If task has terminal state or end point it's called an

• Episodic task

• Continuous task

Self driving cars topic is considered ............ Action space

• Discrete

• Continuous