JAWAHARLAL NEHRU TECHNOLOGICAL UNIVERSITY HYDERABAD B.Tech IV ARTIFICIAL INTELLIGENCE Previous Question Paper (February/March - 2022 - R18)

Agent programs are categorized based on how they map percepts to actions. The four primary types are:

1. Simple Reflex Agents: These operate on a condition-action rule basis, selecting actions based only on the current percept, ignoring the history. They are efficient but limited to fully observable environments.
2. Model-Based Reflex Agents: These maintain an internal state that tracks the part of the world they cannot see. This allows them to handle partially observable environments by updating their internal model of the world.
3. Goal-Based Agents: These agents use goal information to describe desirable situations. They choose actions that help achieve these goals, often involving search and planning to determine the sequence of actions required.
4. Utility-Based Agents: These go beyond goals to consider the quality of the outcome, using a utility function to measure the 'happiness' or efficiency of a state. This allows for decision-making when multiple goals conflict or when there are varying degrees of success.

The A* algorithm is a best-first search algorithm that finds the least-cost path from a start node to a goal node. It uses a evaluation function

For A* to be optimal, the following conditions must be met:

• Admissibility: The heuristic function h(n) must never overestimate the true cost to reach the goal. It must be optimistic, such that
  h(n) ≤ h*(n)
  where h*(n) is the actual minimum cost.
• Consistency (or Monotonicity): For every node n and every successor n' generated by action a, the estimated cost of reaching the goal from n must be no greater than the step cost to n' plus the estimated cost from n' to the goal:
  h(n) ≤ c(n, a, n') + h(n')
  . Consistency is a stronger condition than admissibility and ensures that the total path cost f(n) is non-decreasing along any path.