1
Important Questions and Answers
CP25C05- Artificial Intelligence and Machine Learning
Syllabus
1. Problem Solving and Knowledge Representation: Solving problems by searching –
adversarial search – constraint satisfaction problems – Logical agents – Propositional logic –
First order logic – Forward chaining – Backward chaining – Ontological Representations and
Reasoning Systems.
Activities: Develop a simple Intelligent Agent that navigates a grid world using search
algorithms.
2. Uncertain Knowledge and Reasoning: Overview of uncertainty and basic probability –
Baye’s rule – Bayesian networks – Hidden Markov models – Kalman filters – Utility
Functions – Decision Networks – Sequential decision problems – Game theory
Activities: Develop a simple Intelligent Agent that makes decisions in adversarial conditions
3. Machine Learning Introduction: Types of learning – Hypothesis space – Inductive bias –
Evaluation – Cross validation – Bias variance – Bias Variance Trade off.
Activities: Develop a simple Intelligent Agent that uses logical reasoning to infer hidden
facts or constraints
4. Supervised Learning: Linear Regression – Logistic Regression - Decision trees:
Classification and regression trees – Neural networks, multilayer perceptron – Support vector
machines: linear and non-linear kernel functions – K nearest neighbours – Ensemble
learning: bagging and boosting.
Activities: Behaviour Modelling, Disease Diagnosis and any other similar ones suitable for
supervised learning
5. Unsupervised and Probabilistic Learning: Clustering: partition, hierarchical and density
V.
based clustering – Self organizing maps – Expectation maximization – Gaussian mixture
U
m
models – Principal component analysis – Bayesian learning: Bayes optimal classifier, Naïve
ap
Bayes classifier, Bayesian belief networks.
at
S.No. Description
hy,
Unit-I
Pr
Problem Solving and Knowledge Representation
of
./E
C
N
ot
es
, 2
Basic Concepts
S.
Basic Concept Important Definition / Equation / Notation
No.
Finds a sequence of actions from the initial state to the goal state.
Problem Solving by
1 Search cost: f(n) = g(n) + h(n), where g(n) is path cost and h(n) is
Search
heuristic cost (A* search).
Breadth-First Search Explores nodes level by level. Complete for finite branching; optimal
2
(BFS) when all step costs are equal.
Depth-First Search Explores the deepest node first. Uses less memory but may not find an
3
(DFS) optimal solution.
Used in competitive problems such as games. Minimax equation: V(s)
4 Adversarial Search
= max V(child) for MAX; V(s) = min V(child) for MIN.
Constraint Satisfaction Defined by variables (X), domains (D), and constraints (C). CSP = (X,
5
Problem (CSP) D, C). A solution satisfies all constraints.
Agents use knowledge and inference to select actions. KB = knowledge
6 Logical Agents base; TELL(KB, sentence) adds knowledge; ASK(KB, query) checks
Uses propositions and connectives. NOT: ¬P; AND: P ∧ Q; OR: P ∨
what follows.
7 Propositional Logic
Q; implication: P → Q; equivalence: P ↔ Q.
quantifiers. ∀x P(x) = for all x; ∃x P(x) = there exists an x.
Represents objects, properties, and relations using predicates and
8 First-Order Logic (FOL)
Data-driven inference: starts with known facts and applies rules to
9 Forward Chaining
derive new facts. Rule: P → Q; if P is true, infer Q.
Goal-driven inference: starts with a query and works backward to find
10 Backward Chaining
supporting facts or rules. To prove Q, find a rule P → Q and prove P.
domain. Example: Student ⊆ Person.
Ontological Defines concepts, classes, properties, relationships, and individuals in a
11
Representation
⟹ Q. Used in expert systems and knowledge-based AI.
Derive new knowledge from facts and rules. Modus ponens: P, P → Q
12 Reasoning Systems
Exam tip: Focus on A* search, BFS vs. DFS, minimax, CSP components, propositional logic, FOL
V.
U
quantifiers, forward vs. backward chaining, and modus ponens.
m
ap
Multiple Choice QA
at
h
1. Which search algorithm uses a heuristic function to estimate the cost from the current node to the
y,
goal?
Pr
of
./E
C
N
ot
es
, 3
A) Breadth-First Search (BFS)
B) Depth-First Search (DFS)
C) A Search*
D) Uniform-Cost Search
Answer: C) A Search*
Reason: A* search combines the actual path cost and the estimated remaining cost using the equation:
f(n) = g(n) + h(n)
Where:
g(n) = actual cost from the start node to node n.
h(n) = estimated cost from node n to the goal.
f(n) = estimated total path cost.
2. Which algorithm is commonly used for adversarial search in two-player games such as chess?
A) K-means algorithm
B) Minimax algorithm
C) Linear regression
D) Dijkstra's algorithm
Answer: B) Minimax algorithm
Reason: The Minimax algorithm selects the best move by maximizing the score for one player and
minimizing it for the opponent.
V.
U
Equation:
m
ap
Minimax(n) = max(Value of children) at MAX nodes.
at
h
y,
Minimax(n) = min(Value of children) at MIN nodes.
Pr
of
./E
C
N
ot
es
, 4
Alpha-beta pruning improves Minimax by eliminating branches that cannot affect the final decision.
3. Which technique is used to solve a Constraint Satisfaction Problem (CSP)?
A) Backtracking
B) Gradient descent
C) Linear regression
D) Principal Component Analysis
Answer: A) Backtracking
Reason: Backtracking assigns values to variables one by one and reverses an assignment when a
constraint is violated.
A CSP consists of:
CSP = (X, D, C)
Where:
X = set of variables.
D = domains of possible values.
C = set of constraints.
Example: In map coloring, adjacent regions must have different colors.
4. Which statement correctly describes First-Order Logic (FOL)?
A) It represents only numerical data.
B) It uses objects, predicates, variables and quantifiers.
V.
U
C) It cannot represent relationships between objects.
m
ap
D) It is used only for image processing.
at
hy,
Answer: B) It uses objects, predicates, variables and quantifiers.
Pr
of
./E
C
N
ot
es
Important Questions and Answers
CP25C05- Artificial Intelligence and Machine Learning
Syllabus
1. Problem Solving and Knowledge Representation: Solving problems by searching –
adversarial search – constraint satisfaction problems – Logical agents – Propositional logic –
First order logic – Forward chaining – Backward chaining – Ontological Representations and
Reasoning Systems.
Activities: Develop a simple Intelligent Agent that navigates a grid world using search
algorithms.
2. Uncertain Knowledge and Reasoning: Overview of uncertainty and basic probability –
Baye’s rule – Bayesian networks – Hidden Markov models – Kalman filters – Utility
Functions – Decision Networks – Sequential decision problems – Game theory
Activities: Develop a simple Intelligent Agent that makes decisions in adversarial conditions
3. Machine Learning Introduction: Types of learning – Hypothesis space – Inductive bias –
Evaluation – Cross validation – Bias variance – Bias Variance Trade off.
Activities: Develop a simple Intelligent Agent that uses logical reasoning to infer hidden
facts or constraints
4. Supervised Learning: Linear Regression – Logistic Regression - Decision trees:
Classification and regression trees – Neural networks, multilayer perceptron – Support vector
machines: linear and non-linear kernel functions – K nearest neighbours – Ensemble
learning: bagging and boosting.
Activities: Behaviour Modelling, Disease Diagnosis and any other similar ones suitable for
supervised learning
5. Unsupervised and Probabilistic Learning: Clustering: partition, hierarchical and density
V.
based clustering – Self organizing maps – Expectation maximization – Gaussian mixture
U
m
models – Principal component analysis – Bayesian learning: Bayes optimal classifier, Naïve
ap
Bayes classifier, Bayesian belief networks.
at
S.No. Description
hy,
Unit-I
Pr
Problem Solving and Knowledge Representation
of
./E
C
N
ot
es
, 2
Basic Concepts
S.
Basic Concept Important Definition / Equation / Notation
No.
Finds a sequence of actions from the initial state to the goal state.
Problem Solving by
1 Search cost: f(n) = g(n) + h(n), where g(n) is path cost and h(n) is
Search
heuristic cost (A* search).
Breadth-First Search Explores nodes level by level. Complete for finite branching; optimal
2
(BFS) when all step costs are equal.
Depth-First Search Explores the deepest node first. Uses less memory but may not find an
3
(DFS) optimal solution.
Used in competitive problems such as games. Minimax equation: V(s)
4 Adversarial Search
= max V(child) for MAX; V(s) = min V(child) for MIN.
Constraint Satisfaction Defined by variables (X), domains (D), and constraints (C). CSP = (X,
5
Problem (CSP) D, C). A solution satisfies all constraints.
Agents use knowledge and inference to select actions. KB = knowledge
6 Logical Agents base; TELL(KB, sentence) adds knowledge; ASK(KB, query) checks
Uses propositions and connectives. NOT: ¬P; AND: P ∧ Q; OR: P ∨
what follows.
7 Propositional Logic
Q; implication: P → Q; equivalence: P ↔ Q.
quantifiers. ∀x P(x) = for all x; ∃x P(x) = there exists an x.
Represents objects, properties, and relations using predicates and
8 First-Order Logic (FOL)
Data-driven inference: starts with known facts and applies rules to
9 Forward Chaining
derive new facts. Rule: P → Q; if P is true, infer Q.
Goal-driven inference: starts with a query and works backward to find
10 Backward Chaining
supporting facts or rules. To prove Q, find a rule P → Q and prove P.
domain. Example: Student ⊆ Person.
Ontological Defines concepts, classes, properties, relationships, and individuals in a
11
Representation
⟹ Q. Used in expert systems and knowledge-based AI.
Derive new knowledge from facts and rules. Modus ponens: P, P → Q
12 Reasoning Systems
Exam tip: Focus on A* search, BFS vs. DFS, minimax, CSP components, propositional logic, FOL
V.
U
quantifiers, forward vs. backward chaining, and modus ponens.
m
ap
Multiple Choice QA
at
h
1. Which search algorithm uses a heuristic function to estimate the cost from the current node to the
y,
goal?
Pr
of
./E
C
N
ot
es
, 3
A) Breadth-First Search (BFS)
B) Depth-First Search (DFS)
C) A Search*
D) Uniform-Cost Search
Answer: C) A Search*
Reason: A* search combines the actual path cost and the estimated remaining cost using the equation:
f(n) = g(n) + h(n)
Where:
g(n) = actual cost from the start node to node n.
h(n) = estimated cost from node n to the goal.
f(n) = estimated total path cost.
2. Which algorithm is commonly used for adversarial search in two-player games such as chess?
A) K-means algorithm
B) Minimax algorithm
C) Linear regression
D) Dijkstra's algorithm
Answer: B) Minimax algorithm
Reason: The Minimax algorithm selects the best move by maximizing the score for one player and
minimizing it for the opponent.
V.
U
Equation:
m
ap
Minimax(n) = max(Value of children) at MAX nodes.
at
h
y,
Minimax(n) = min(Value of children) at MIN nodes.
Pr
of
./E
C
N
ot
es
, 4
Alpha-beta pruning improves Minimax by eliminating branches that cannot affect the final decision.
3. Which technique is used to solve a Constraint Satisfaction Problem (CSP)?
A) Backtracking
B) Gradient descent
C) Linear regression
D) Principal Component Analysis
Answer: A) Backtracking
Reason: Backtracking assigns values to variables one by one and reverses an assignment when a
constraint is violated.
A CSP consists of:
CSP = (X, D, C)
Where:
X = set of variables.
D = domains of possible values.
C = set of constraints.
Example: In map coloring, adjacent regions must have different colors.
4. Which statement correctly describes First-Order Logic (FOL)?
A) It represents only numerical data.
B) It uses objects, predicates, variables and quantifiers.
V.
U
C) It cannot represent relationships between objects.
m
ap
D) It is used only for image processing.
at
hy,
Answer: B) It uses objects, predicates, variables and quantifiers.
Pr
of
./E
C
N
ot
es
