Question Paper Code : 243307

ARUNAI ENGINEERING COLLEGE(AUTONOMOUS)

B.E/B.Tech.DEGREE EXAMINATION,Nov/Dec-2024
ARTIFICIAL INTELLIGENCE AND DATA SCIENCE
II Year/III Semester
AL3391-ARTIFICIAL INTELLIGENCE
(Regulations 2021)
ANSWER KEY
PART-A (10x2=20 Marks)

1. Explain Agent and Environment

An agent can be anything that perceive its environment through sensors and act upon that
environment through actuators.
An environment is everything in the world which surrounds the agent, but it is not a part of
an agent itself.

2. Describe how will you Measure problem-solving performance ?

We can evaluate an algorithm’s performance in four ways:
Completeness : Is the algorithm guaranteed to find a solution when there is one?
Optimality : Does the strategy find the optimal solution?
Time complexity : How long does it take to find a solution?
Space complexity : How much memory is needed to perform the search?
3. Define heuristic.Why are heuristic crucial for the efficient design of an expert system.
A heuristic is a rule of thumb or an approach that helps in problem-solving by providing a
practical, efficient solution that may not be optimal but works well in most situations.
Heuristics are crucial in the efficient design of expert systems because:
They reduce the search space by guiding the system towards likely solutions, improving
efficiency.
They allow the system to make decisions or recommendations in complex, real-world scenarios
where complete information may not be available.

4. Explain what is optimal solution?

If a solution found for an algorithm is guaranteed to be the best solution (lowest path
cost) among all other solutions, then such a solution for is said to be an optimal solution.

5. What is your understanding on stochastic games?

Stochastic Games are multi-agent decision-making models where players' actions and
random events determine the transitions between states. Players aim to maximize cumulative
rewards over time by considering both immediate payoffs and future state probabilities.
They are used in areas like economics, AI, and robotics to model competition, collaboration, and
decision-making in uncertain environments.

6. Define Backtracking

1
 Backtracking is a problem-solving technique used in algorithms where a systematic trial-
and-error approach is applied to explore all possible solutions. It involves exploring a path, and if
it leads to a dead end or an invalid solution, the algorithm backtracks to the previous step to try a
different path.
It is commonly used in problems like solving mazes, puzzles (e.g., Sudoku), and constraint
satisfaction problems.

7. Specify the basic components of propositional logic?

Propositions: Simple declarative statements that can be either true or false (e.g., "It is
raining").
Logical Connectives: Symbols used to combine propositions, such as:
a. AND (∧)
b. OR (∨)
c. NOT (¬)
d. IMPLIES (→)
e. BICONDITIONAL (↔)
Truth Values: The truth or falsity of propositions, represented as True (T) or False (F).
Syntax: Rules that define how propositions and connectives can be combined to form
valid logical expressions.
Semantics: The interpretation that assigns truth values to propositions and determines the
truth of compound expressions.

8. Define First order Logic.

First-Order Logic (FOL), also known as Predicate Logic, is a formal system used to
represent and reason about relationships, objects, and their properties. It extends propositional
logic by including:
Quantifiers:Universal Quantifier (∀) & Existential Quantifier (∃)
Predicates,Constants,Variables & Functions
FOL is more expressive than propositional logic and is used in areas like knowledge
representation, artificial intelligence, and automated reasoning.

9. Define naive Bayes model.

The Naive Bayes Model is a probabilistic classification algorithm based on Bayes'
theorem. It assumes that the features (attributes) are independent of each other given the class
label, which is referred to as the "naive" assumption.

10. Write about Uncertainty.

2
 Uncertainty in artificial intelligence refers to the lack of complete information about a
system, environment, or situation, leading to ambiguity in decision-making or predictions. It
arises from:
1. Incomplete Data: Missing or unavailable information.
2. Noisy Data: Errors or inconsistencies in data.
3. Randomness: Inherent variability in systems or processes.
Techniques like probability theory, fuzzy logic, and Bayesian models are used to handle
uncertainty in AI systems.
PART-B (5x13=65 Marks)
11. A) Define an agent.Explain the four basic agent and the principles of intelligent systems
with example. (13)

An agent can be anything that perceive its environment through sensors and act upon that
environment through actuators.

Agents can be grouped into five classes based on their degree of perceived intelligence and
capability. All these agents can improve their performance and generate better action over the
time. These are given below:
(ANY FOUR)
• Simple Reflex Agent
• Model-based reflex agent (3)
• Goal-based agents
• Utility-based agent
• Learning agent
Simple Reflex Agent (5)

Model-based reflex agent (8)

Goal-based agent

3
Utility-based agent (13)

Learning agent

(OR)
11.B (i) What are informed search techniques? Expalin any one serach techiques. (8)
In the informed search we will discuss two main algorithms which are given below:
ANY ONE
o Best First Search Algorithm(Greedy search)
o A* Search Algorithm

Best-first Search Algorithm (Greedy Search): (2)

Greedy best-first search algorithm always selects the path which appears best at
that moment. It is the combination of depth-first search and breadth-first search
algorithms. It uses the heuristic function and search.we expand the node which is
closest to the goal node and the closest cost is estimated by heuristic function, i.e.
f(n)= g(n).
h(n)= estimated cost from node n to the goal.
Best first search algorithm: (4)

4
 o Step 1: Place the starting node into the OPEN list.
o Step 2: If the OPEN list is empty, Stop and return failure.
o Step 3: Remove the node n, from the OPEN list which has the lowest value
of h(n), and places it in the CLOSED list.
o Step 4: Expand the node n, and generate the successors of node n.
o Step 5: Check each successor of node n, and find whether any node is a
goal node or not. If any successor node is goal node, then return success
and terminate the search, else proceed to Step 6.
o Step 6: For each successor node, algorithm checks for evaluation function
f(n), and then check if the node has been in either OPEN or CLOSED list.
If
Example:Consider the below search problem, and we will traverse it using greedy best-first
search. At each iteration, each node is expanded using evaluation function f(n)=h(n) , which

is given in the below table.

In this search example, we are using two lists which are OPEN and CLOSED Lists. (8)
Hence the final solution path will be: S----> B----->F- - -> G

Time Complexity: O(bm).

Space Complexity: O(bm). Where, m is the maximum depth of the search space.
Complete: Greedy best-first search is also incomplete, even if the given state space is finite.
Optimal: Greedy best first search algorithm is not optimal.
11.B (ii)List the Basic Kinds of Intelligent Agents (5)
1. Simple Reflex Agents: These agents act solely based on the current percept without
considering the history or future consequences. (1)
2. Model-Based Reflex Agents: These agents maintain an internal model of the
environment to handle partially observable environments. (2)
3. Goal-Based Agents: These agents make decisions based on specific goals they are
designed to achieve. (3)
4. Utility-Based Agents: These agents aim to maximize a utility function, which quantifies
how desirable a state is. (4)
5. Learning Agents: These agents improve their performance over time by learning from
experiences and the environment. (5)

5
12.A) Discuss on Online Search Agents that uses depth-first exploration. (13)
Depth-First Exploration in Online Search Agents
1. Mechanism: (4)
o The agent explores as deep as possible into the search tree before backtracking.

o It selects an unexplored action at each state and moves forward.

o If it reaches a dead end or an already visited state, it backtracks to the previous

state and tries another action.
2. Features: (8)
o Incremental Exploration: Only explores one branch at a time without building a
complete map of the environment.
o Memory-Efficient: Requires minimal memory as it only needs to track the
current path (stack structure).
o Simple Implementation: Easy to implement and suitable for environments with
limited resources.

Applications: (10)
 Maze navigation
 Real-time robot pathfinding
 Games where exploration and immediate decisions are crucial
Advantages:
 Low Memory Usage: Unlike breadth-first exploration, it doesn’t store all nodes at the
same level.
 Quick Solution Discovery: Can quickly find a solution if it lies along the first explored
branch.
Disadvantages: (13)
 Suboptimal Solutions: May not find the shortest or optimal path.
 Repeated States: Without proper mechanisms, the agent may revisit states or get stuck in
loops. (OR)

12. B) What is heuristic search techinque in AI? How does heuristic search works? Explain
its advantages and disadvantages. (13)

A heuristic is a rule of thumb or an approach that helps in problem-solving by providing a

practical, efficient solution that may not be optimal but works well in most situations. (4)
Some of the real-life examples of heuristics that people use as a way to solve a problem:
o Common sense: It is a heuristic that is used to solve a problem based on
the observation of an individual.
o Rule of thumb: In heuristics, we also use a term rule of thumb. This
heuristic allows an individual to make an approximation without doing an
exhaustive search.

6
 o Working backward: It lets an individual solve a problem by assuming that
the problem is already being solved by them and working backward in their
minds to see how much a solution has been reached.
o Availability heuristic: It allows a person to judge a situation based on the
examples of similar situations that come to mind.
o Familiarity heuristic: It allows a person to approach a problem on the fact
that an individual is familiar with the same situation, so one should act
similarly as he/she acted in the same situation before.
o Educated guess: It allows a person to reach a conclusion without doing an
exhaustive search.
Heuristic search techinque in AI
A* Search Algorithm
8- Puzzle problem
Hill Climbing
8-Puzzle problem (8)

There are several pathways in a search tree to reach the goal node from the current node. The
selection of a good heuristic function matters certainly.

ADV: Complete and optimal,used to solve complex problems (13)

DISADV: Requires more memory space

7
13.A) What are Partially observable games? How are they solved in a deterministic
environment (13)

Partially observable games are games in which players do not have complete knowledge
of the game state due to hidden or incomplete information. This partial observability can arise
from hidden states (e.g., cards in an opponent's hand) or imperfect sensors (e.g., in robotics).
Such games are common in many real-world scenarios, such as poker, battleships, or military
strategy. (4)
In deterministic environments, the outcome of every action is predictable, but players must still
account for their uncertainty about the game state caused by their limited information. (6)
Solving Partially Observable Games in a Deterministic Environment
To solve these games, players typically use strategies that incorporate reasoning about both the
game dynamics and their uncertainty. Here are some approaches:
Minimax Search with Incomplete Information
The minimax algorithm can be adapted for partial observability by considering the belief state
instead of the exact state:
 The player assumes the worst-case scenario for any hidden information (conservative
approach).
 Actions are chosen to minimize the maximum possible loss. (10)
Heuristic Search
When exact solutions are infeasible, heuristics can be used to guide the search:
 Define heuristic evaluations based on available information.
 Prune actions or states unlikely to yield optimal results.
Example (13)
In a simplified poker game:
 You know your cards but not your opponent's cards.
 You maintain a belief about their hand based on observed actions (e.g., betting or
folding).
 You choose actions (e.g., betting or folding) based on maximizing expected utility given
your belief state. (OR)

13.B) Explain the constraint satisfaction problem and the variations of constraint
satisfaction problem with example ? (13)
A Constraint Satisfaction Problem (CSP) is a type of mathematical problem characterized by:
1. Variables: A set of variables X={X1,X2,...,Xn}. (6)
2. Domains: Each variable Xi has a finite set of possible values called its domain D
3. Constraints: A set of rules that specify allowable combinations of values for a subset of
variables.
CSP Problems (8)
Graph Coloring
Crypt Arthimetic problem
N-Queen problem(Sudoku)

8
Given a crypt-arithmetic problem, i.e., S E N D + M O R E = M O N E Y (10)

(13)

14.A(i) Brief on the Concept of Resolution and explain the propositional resolution
Algorithm (6)

9
 Resolution is a fundamental rule of inference used in Artificial Intelligence (AI) for
automated theorem proving and reasoning in propositional and first-order logic. (2)
Propositional Resolution Algorithm:
The Propositional Resolution Algorithm is a systematic method used for checking the
satisfiability of a propositional logic formula. It works as follows: (6)
Step 1: Convert the formula into CNF.
Step 2: Negate the conclusion.
Step 3: Apply the resolution rule.
Step 4: Repeat until

14.A(ii) Prove the following axioms using the resolution Algorithm. (7)

(OR)
14.B) Discuss the Knowledge Engineering Process with proper illustration. Depict the
concept of forward chaining. (13)
The knowledge engineering process involves creating systems that use expert knowledge to
solve problems or make decisions. It is the backbone of artificial intelligence (AI) systems,
especially in areas like expert systems, decision support systems, and rule-based systems.
Steps in the Knowledge Engineering Process (4)
1. Identify the Problem Domain
o Understand the domain where the system will be applied.
2. Gather Knowledge
o Collect information from domain experts, textbooks, databases, or research
papers.
3. Represent Knowledge (6)

10
 o Choose an appropriate representation format for knowledge (e.g., rules, frames,
semantic networks).
4. Build the Knowledge Base
o Create a repository of the knowledge using the chosen representation.
5. Develop the Inference Engine
o Create a mechanism to apply the knowledge base to solve problems.
6. Integrate the User Interface (10)
o Build an interface that allows users to interact with the system.
7. Test the System
o Validate the system using test cases to ensure accuracy and reliability.
8. Maintain and Update
o Periodically update the knowledge base as new information becomes available.
Concept of Forward Chaining (13)
Forward chaining is an inference technique used in rule-based systems. It starts with the
known facts and applies rules to infer new facts until a conclusion is reached.

15. A(i) Discuss Bayesian belief network for classification using rain prediction dataset.(8)

A Bayesian Belief Network (BBN) is a probabilistic graphical model that represents the
conditional dependencies among a set of variables. It consists of: (4)
1. Nodes: Representing variables (e.g., weather conditions, rain).
2. Edges: Representing conditional dependencies.
3. Conditional Probability Tables (CPTs): Quantifying the relationships between
variables.
Rain Prediction Dataset (8)
Consider a simple dataset with the following features:
Temperature Humidity Windy Rain
Hot High True Yes

Mild High False No

Cool Normal True Yes

Hot Low False No

Mild Normal True Yes

Temperature, Humidity, and Windy are features (inputs).Rain is the target variable (output).
15.A(ii) Discuss the casual network with example. (5)
A causal network is a graphical model that represents cause-and-effect relationships between
variables. It is a type of Bayesian Belief Network (BBN) but explicitly focuses on causation
rather than correlation. (2)
The network is represented by: (5)

11
 1. Nodes: Variables or events.
2. Edges: Directed edges that indicate causal relationships (e.g., "A causes B").
3. Conditional Probability Tables (CPTs): Quantify the strength of the causal
relationship. (OR)
15.B(i) Discuss the uncertain knowledge with example. (8)
Uncertain knowledge refers to situations where we do not have complete or precise
information about the state of the world. In such cases, we rely on probabilistic reasoning or
other techniques to manage uncertainty and make decisions based on incomplete or ambiguous
information. (2)
Example: Diagnosing a Disease
Consider a doctor diagnosing a patient based on uncertain knowledge.
Scenario: (4)
 A patient has symptoms like fever and fatigue.
 The doctor suspects it might be malaria or flu.
Uncertain Knowledge: (8)
1. The presence of fever is 80% likely in malaria cases.
2. Fatigue is 60% likely in flu cases.
3. There is a 20% prior probability of malaria in the area and a 30% prior probability of
flu.
15.B(ii) Briefly explain the probalistic reasoning. (5)
Probabilistic reasoning is a method of reasoning and decision-making under uncertainty. It
uses the principles of probability theory to model, represent, and infer knowledge when the
information available is incomplete, ambiguous, or uncertain. (2)
Steps in Probabilistic Reasoning
1. Define Variables2. Determine Relationships
2. Assign Probabilities 4. Inference. (5)

PART- C (1x15=15 Marks)

16. A) How does Alpha beta search algorithm differ from Minmax algorithm.Analyse and
explain the significance of knowledge representation? Mention the desirable properties of
knowledge representation. (15)

Difference Between Alpha-Beta Search Algorithm and Minimax Algorithm

The Minimax algorithm and the Alpha-Beta pruning algorithm are both used in decision-making
for adversarial games (e.g., chess, checkers). The key difference lies in efficiency and the ability
of Alpha-Beta to prune unnecessary nodes. (5)
1. Minimax Algorithm
 A recursive algorithm for finding the optimal move in a two-player game.
 It assumes:
o One player maximizes the score (Max).
o The opponent minimizes the score (Min).
2. Alpha-Beta Pruning
 It maintains two parameters:
o Alpha: The best value that the maximizing player can guarantee.

12
 Beta: The best value that the minimizing player can guarantee.
o
 Pruning: It stops evaluating branches of the tree that cannot affect the final decision,
improving efficiency.
EXAMPLE: (10)

Significance of Knowledge Representation (13)

Knowledge representation (KR) is the way information and facts about the world are encoded so
that a machine can reason, learn, and make decisions.
Desirable Properties of Knowledge Representation (15)
1. Representational Adequacy
2. Inferential Adequacy
3. Inferential Efficiency
4. Acquisitional Efficiency
5. Modularity
6. Expressiveness
7. Robustness (OR)

16. B) Explain brief discussion of Hill Climbing algorithm with its suitable example.(15)
The Hill Climbing Algorithm is a local search optimization technique used to find a solution
to a problem by iteratively improving a single solution (4)
Steps in the Hill Climbing Algorithm (6)
1. Start with an initial solution.
2. Evaluate the current state using an objective function.
3. Move to the best neighboring state.
o If no neighbor improves the state, terminate (local optimum).
4. Repeat steps 2-3 until a goal is reached or no better neighbors exist.
Variants of Hill Climbing
1. Simple Hill Climbing: (12)
o Evaluates one neighbor at a time and moves if improvement is found.
2. Steepest-Ascent Hill Climbing:

13
 oEvaluates all neighbors and selects the one with the highest improvement.
3. Stochastic Hill Climbing:
o Selects a random neighbor and moves based on probability of improvement.

Example: Hill Climbing in the Traveling Salesman Problem (15)

Find the shortest path to visit a set of cities and return to the starting point.
Steps:
1. Initial Solution: Start with a random path between cities (e.g., A → B → C → D → A).
2. Objective Function: Minimize the total distance of the path.
3. Neighboring States: Swap the positions of two cities in the current path (e.g., A → C →
B → D → A).
4. Improvement: Check if the new path has a shorter distance. If yes, move to this state;
otherwise, try another neighbor.
5. Repeat until no shorter path is found.

Prepared By Verified By Approved By

14