JNTU B.Tech III ARTIFICIAL INTELLIGENCE Previous Question Paper (January/February - 2021 - R16)

An Artificial Intelligence technique is a method that exploits knowledge that is represented in such a way that the knowledge itself facilitates the solution to a problem. To be considered an effective AI technique, the method must be general enough to handle a variety of situations, yet specific enough to provide a clear path to a solution. Key characteristics include the ability to represent large amounts of knowledge, the capacity to manipulate this knowledge to create new information, and the flexibility to operate even when the input data is incomplete or ambiguous. These techniques often involve search algorithms, heuristic functions, and logical inference mechanisms to navigate complex state spaces efficiently.

In the design of an intelligent agent, goal formulation must precede problem formulation because the goal defines the desired end state, which serves as the essential reference point for determining what actions are necessary. Without a clear goal, an agent cannot distinguish between relevant and irrelevant information, nor can it evaluate the success of its actions.

The dependency exists for the following reasons:

• Defining the Objective: Goal formulation establishes the criteria for success. Problem formulation then focuses on identifying the states and actions required to reach that specific objective.
• Filtering the State Space: The problem formulation process involves defining the initial state, the set of possible actions, and the transition model. If the goal is not defined first, the agent might consider an infinite or overly complex set of irrelevant states, leading to computational inefficiency.
• Determining Constraints: The goal dictates the constraints of the problem. For example, if the goal is to reach a destination in the shortest time, the problem formulation must include time-based costs for each action. If the goal is simply to reach the destination, those costs may be ignored.
• Heuristic Guidance: Effective problem-solving often relies on heuristics. These heuristics are derived from the goal state; therefore, the goal must be known to construct an informed search strategy that guides the agent toward the target.

Ultimately, goal formulation provides the "what" (the target), while problem formulation provides the "how" (the path), ensuring that the agent's computational resources are directed toward a meaningful and achievable outcome.