Agent Environment in Artificial Intelligence

Before, we start to learn about the Agent Environment, one important aspect about the agent is rationality  and it should be covered first. Whenever, the agent perform any task that task should be correct as per sequence of information provided known as agent's rationality. We can consider it as good behavior of the agent. 

Good behavior or rationality can be achieved by following- 

1. Performance Measurement is the criteria of the success 
2. Agent is having prior knowledge of its environment
3. Sequence of action performed by agent
4. Agent's perception of the sequence 

Now lets talk about the environment or we can say task environment of any agent. We can consider the environment as a  "Problem" and consider the Agent as a "Solution". So, we have variety of environment and agents.

For better understanding lets see some examples:

1. Agent- Medical Diagnostic System
    Environment- Hospital, Doctors, Nurses

2. Agent- English Teacher
    Environment- Class, Student, Class Test

Type of Task Environment in Artificial Intelligence

1. Fully Observable and Partial Observable Environment: As you know agents sensor provide the information about the agents environment. If agent's sensor is providing all the information of the state of environment are known as Fully Observable Environment. If agent's sensor is providing partial information (due to noisy sensor or missing environment information) of the state of environment known as Partial Observable Environment.

2. Single Agent and Multi-Agent Environment: If the single agent capable enough to perform a task known as Single Agent Environment. For example- An agent that can solve SUDOKU. 
Multi-Agent Environment is used multiple agents to perform the task. Fro example- to play the CHESS we require two agents.

3. Deterministic and Stochastic Environment: If the agents determine the next state of the environment through current state of the environment known as Deterministic Environment else Stochastic Environment.

4. Episodic and Sequential Environment: If the agent divided entire task in the individual task and perform single action in the individual task that are not depended to previous task know as Episodic Environment. Whereas, in Sequential Environment current task could affect all future  task.

5. Static and Dynamic Environment: If the environment of the agent is not changing every time can known as static environment else known as dynamic Environment.

6. Discrete and Continuous Environment: To understand the Discrete and Continuous Environment , chess is the example of discrete environment where agents take discrete steps. Driver agent is the example of continuous environment.

7. Known and Unknown Environment: If agent have all prior knowledge about the environment and its outcome is  called as known environment. If the agent don't have prior knowledge of environment than agent will have to collect all the information about the state of environment for better decisions.