Home »
Machine Learning/Artificial Intelligence
Classification of Environment in Artificial Intelligence
Artificial Intelligence | Types of environment: In this tutorial, we are going to learn about the different types of environment present for an agent (Artificial Intelligence Based Agent). We will also learn about the criteria of classification upon which these certain types of environment are defined.
By Monika Sharma Last updated : April 12, 2023
The environment is the part of the universe which surrounds the intelligent system (Agent). As per the definition of artificial intelligence, the environment, and the agent together make up the Universe. But there are certain classifications of the environment by which we can define the type of environment the agent can handle or works.
The environment can be classified as follows:
1. Accessible and Inaccessible
The first classification of environment deals with the accessible region of the environment by the agent. Although everything present around the agent is termed as environment, still the whole part may not be in the agent’s reach and hence can be inaccessible by the agent. Therefore, we can further differentiate the environment on the basis whether it is accessible by the agent or not.
2. Deterministic or Non-Deterministic
While dealing with the environment, the agent may not always be absolutely sure about the observations and perceptions it makes from it. Therefore, the environment can be divided on the basis that whether the agent can be fully determined about the decisions it makes from the environment or is non-deterministic about it.
3. Static or Dynamic
The environment for the agent may not always be the same as it was initially. It may change with time and conditions. So, the environment can also be differentiated on the basis e=whether it is static or dynamic. If the environment does not change its state with the passage of time, then it is termed as static. If there are changes in the environment with time, and its state does not remain constant throughout, then the environment is described as dynamic.
4. Discrete or continuous
Sometimes, the observations made by the agent cannot be taken continuously. This totally depends on the type of environment it works in. In this type of classification, the environment is defined either as discrete or continuous. If the agent can perceive and make observations from the environment continuously without any lag, then the environment is termed as continuous. If the agent observes the environment only for certain instants and the observations are not continuous, then the environment is termed as discrete.
5. Observable or Partially-observable
The environment in which the agent works may not always be completely observable by the agent. Some part of the environment may be there that is out of the reach of the agent and the agent is not able to include that portion of the environment in its perception. So, the environment is categorized as observable and partially observable. In the observable environment, the agent can make observations from its complete surroundings, but in a partially observable environment, some parts of its surroundings are there which exist but are not observable by the environment.