Department of Computer Science and Engineering
(Artificial Intelligence and Machine Learning)
Continuous Internal Evaluation Test I - ODD Semester 2025 - 26
Course Title: Advanced AI and ML Course Code: AM722I1A
Scheme and Solution
Q. No. Questions Marks
Apply the Minimax algorithm to find the best move for MAX and indicate the minimax
value at the root. Show all steps.
a
The algorithm works by evaluating the game tree from the bottom up. At the MIN level,
the node takes the minimum value of its children. At the MAX level, the node takes the
maximum value of its children. This continues all the way up to the root.
8
1
The root node is a MAX node. It takes the maximum of its children:
Root MAX node: max(2,0)=2
b Explain the structure of goal-based and utility-based agents with neat labeled diagrams
and their pseudocode.
Goal-based agents 8
A goal-based agent decides actions not just based on the current state but also on a
goal—a desired outcome or final state it aims to achieve.
, It uses both the model of the world (like a model-based reflex agent) and goal
information to plan and act effectively.
Example: “Brake lights mean car in front is slowing. If I don’t brake, I’ll crash. My goal is
safe driving. Therefore, brake.”
Pseudocode:
function GOAL-BASED-AGENT(percept) returns an action
persistent: state, the agent’s current conception of the world state
model, a description of how the world evolves
goals, a set of desirable states
plan, a sequence of actions to achieve a goal, initially empty
state ← UPDATE-STATE(state, percept, model)
if plan is empty then
goal ← SELECT-GOAL(goals, state)
problem ← FORMULATE-PROBLEM(state, goal, model)
plan ← SEARCH(problem)
action ← FIRST(plan)
plan ← REMAINING(plan)
return action
Utility-based agents
A utility-based agent goes beyond goals.
It uses a utility function to evaluate different states and actions, choosing the one that
maximizes expected utility.
Example: Get passenger to the destination safely, quickly, and cheaply by weighing trade-
offs (avoid traffic jams, save fuel, minimize accident risk).
(Artificial Intelligence and Machine Learning)
Continuous Internal Evaluation Test I - ODD Semester 2025 - 26
Course Title: Advanced AI and ML Course Code: AM722I1A
Scheme and Solution
Q. No. Questions Marks
Apply the Minimax algorithm to find the best move for MAX and indicate the minimax
value at the root. Show all steps.
a
The algorithm works by evaluating the game tree from the bottom up. At the MIN level,
the node takes the minimum value of its children. At the MAX level, the node takes the
maximum value of its children. This continues all the way up to the root.
8
1
The root node is a MAX node. It takes the maximum of its children:
Root MAX node: max(2,0)=2
b Explain the structure of goal-based and utility-based agents with neat labeled diagrams
and their pseudocode.
Goal-based agents 8
A goal-based agent decides actions not just based on the current state but also on a
goal—a desired outcome or final state it aims to achieve.
, It uses both the model of the world (like a model-based reflex agent) and goal
information to plan and act effectively.
Example: “Brake lights mean car in front is slowing. If I don’t brake, I’ll crash. My goal is
safe driving. Therefore, brake.”
Pseudocode:
function GOAL-BASED-AGENT(percept) returns an action
persistent: state, the agent’s current conception of the world state
model, a description of how the world evolves
goals, a set of desirable states
plan, a sequence of actions to achieve a goal, initially empty
state ← UPDATE-STATE(state, percept, model)
if plan is empty then
goal ← SELECT-GOAL(goals, state)
problem ← FORMULATE-PROBLEM(state, goal, model)
plan ← SEARCH(problem)
action ← FIRST(plan)
plan ← REMAINING(plan)
return action
Utility-based agents
A utility-based agent goes beyond goals.
It uses a utility function to evaluate different states and actions, choosing the one that
maximizes expected utility.
Example: Get passenger to the destination safely, quickly, and cheaply by weighing trade-
offs (avoid traffic jams, save fuel, minimize accident risk).
