1. Simple Reflex Agents:

Simple Reflex Agents forgets about the past history of percepts and works upon only on present percept. The Agent function is based on condition action rule. Condition Action Rule is the rule that maps the state/condition to actions. If the condition is true then the action will take otherwise not. This Agents function Succeed only when the entire Environment is fully observable. For a simple Reflex Agents when the Environment is partially observable then the problem of infinite loops is unavoidable. It may be possible to avoid infinite loops if the Agent can randomize its action. There are some problems with Simple Reflex Agents:

• • Very limited Intelligence
• • No knowledge on non-perceptual parts of the state
• • Usually too big to generate and store
• • If there occurs any change in the Environment, then the collection of rules needs to be updated.

2. Model based Reflex Agents:

A model-based Agent can handle partially observable environments by the use of model about the world. The Agent has to find a rule whose condition matches the current situation of the Environment. It always keep track of internal state which is adjusted by each percept and that depends on the percept history. The current state is stored inside the agent which maintains some kind of structure describing the part of the Environment which can’t be seen. Updating the state requires about the agent which maintains some kind of information about that: How the world involves independently from the Agent, and how Agents action effect the world.

3. Goal based Agents:

This kind of Agents have a particular goal to achieve so the Agents always tries to reduce the distance towards their goal and takes decision based on how far they are currently form the goal. This allows an Agent to select among multiple paths/ possibilities selecting the one to reach the goal state. The knowledge that supports to take decision can be modified and that is represented explicitly which makes the Agent more flexible. They usually represent planning and search. The Goal-based Agents behaviour can easily change according to the situation or can be modified.

4. Utility based Agents:

In utility-based Agents Utility describes the state of happiness of the Agent. Sometimes only to reach ndestination is not enough, Agents have to achieve the goal in a safer, cheaper and quicker manner to achieve the goal, for that reason Utility-based Agents choose the Actions that maximized the expected utility to chose the best path to reach the goal. A utility function maps a state into real number which describes the associate degree of happiness.

5. Learning Agents:

Learning Agents have the ability to learn from the past experience and it starts to act with the basic knowledge and then it is able to act and adapt the knowledges from the Environment automatically through the learning. It has basically 4 components, these are: