P1 · 82465 (May 2025)
P2 · 65192 (Dec 2024)
P3 · 54516
P4 · 26341
P5 · 38793 (Dec 2023)
P6 · 12281 (Dec 2022)
P7 · 93660 (Summer 2022 MCQ)
MathJax enabled
01
Introduction to Artificial Intelligence
1.1 Introduction · History of AI · Intelligent Systems: Categorization · Components of AI Program · Foundations of AI · Sub-areas · Applications · Current Trends
4 HRS
Define the term AI. List down all components of an AI Program.
Describe different categories / categorization of AI. (P6: Explain the categorization of Intelligent System.)
What do you mean by Total Turing Test? Explain.
Illustrate / Describe the application areas of AI in Robotics. How AI will help in Robotics application?
The computer program that simulates the thought process of humans is known as:
A) Expert reason B) Personal information C) Expert system D) Human logic
The father of AI is:
A) Alan Turing B) John McCarthy C) Russel Stuart D) Andrew Ng
__________ is the field that investigates the mechanics of human intelligence.
A) Sociology B) Nurology C) Cognitive science D) Psychology
02
Intelligent Agents
2.1 Agents & Environments · Rationality · Nature of Environment · Structure of Agents · Types of Agents · Learning Agent
2.2 Solving Problems by Searching: Problem Solving Agent · Formulating Problems · Example Problems (Water Jug, 8-Puzzle, Autonomous Taxi)
2.2 Solving Problems by Searching: Problem Solving Agent · Formulating Problems · Example Problems (Water Jug, 8-Puzzle, Autonomous Taxi)
4 HRS
2.1 · Types of Agents & Environments
Describe / Explain various properties of task environment (types of environments) of AI agents with suitable examples.
Describe different types of environment of AI agents.
List down all agent types. Explain each with block diagram.
Explain the Learning Agent with a suitable block diagram.
Explain Utility-based Agent architecture with diagram. (P4: Explain Goal-based Agent with a block diagram.)
In which agent does the problem generator is present?
A) Learning agent B) Simple-reflex agent C) Goal based agent D) Utility based agent
2.2 · PEAS Description & Problem Formulation
Give PEAS descriptors for the following agents:
P1: Vacuum cleaner
P3: Automobile Driver Agent (also give state space description)
P4: Medical diagnosis system & Part picking robot
P5: Part picking robot using PEAS properties
P3: Automobile Driver Agent (also give state space description)
P4: Medical diagnosis system & Part picking robot
P5: Part picking robot using PEAS properties
Task of cleaning house has been assigned to a vacuum cleaner robot. Initial location is unknown. Robot has to execute appropriate actions to clean house. Identify the appropriate type of agent, applicable task environment, and PEAS parameters.
Explain PEAS descriptors, also state PEAS description for an online English tutor.
Formulate the problem. Choose a precise formulation and identify the initial state, goal test, successor function, and cost function for the following:
P2: Autonomous Taxi driver
P5: 8-puzzle problem
P6: A 3-foot tall monkey in a room where bananas are suspended from an 8-foot ceiling, with two 3-foot high movable crates.
P5: 8-puzzle problem
P6: A 3-foot tall monkey in a room where bananas are suspended from an 8-foot ceiling, with two 3-foot high movable crates.
Enumerate the Classical "Water Jug Problem". Describe the state space for this problem and also give the solution.
What is formulation of a problem? Formulate the Wumpus World problem in terms of: initial state, actions, successor function, goal test, path cost.
What do you mean by state space representation? Explain with example the necessity of it.
03
Problem Solving
3.1 Uninformed Search: BFS · DFS · Depth Limited · DFID Informed Search: Greedy Best-First · A* · Memory Bounded
3.2 Local Search & Optimization: Hill Climbing · Simulated Annealing · Genetic Algorithms
3.3 Adversarial Search: Game Playing · Min-Max Search · Alpha-Beta Pruning
3.2 Local Search & Optimization: Hill Climbing · Simulated Annealing · Genetic Algorithms
3.3 Adversarial Search: Game Playing · Min-Max Search · Alpha-Beta Pruning
10 HRS
3.1 · Uninformed Search: BFS, DFS, Depth Limited, DFID
Explain Depth Limit Search and Depth First Iterative Deepening (DFID) search.
Give difference between / Explain in detail Informed Search vs Uninformed Search algorithms.
3.1 · Informed Search: Greedy Best-First, A*
Explain A* search algorithm with a suitable example. (P5: Is A* informed or uninformed search — Justify.)
Key formula: \(f(n) = g(n) + h(n)\), where \(g(n)\) = cost so far, \(h(n)\) = heuristic estimate
Apply Greedy Best-First Search. At each iteration, expand using \(f(n) = h(n)\):
h(S)=10, h(A)=10, h(D)=8, h(B)=6, h(E)=6.5, h(C)=4, h(F)=3, h(G)=0
S is start state, G is goal state.
Graph edges: S→A(3), S→D(4), A→B(4), A→C(4), D→E(2), D→B(5), B→E(−), C→G(3), E→F(4), F→G(−)
Perform A* search to find the shortest distance path from S to any of the goal nodes G1, G2, G3:
| Edge | Cost | Edge | Cost | Edge | Cost |
|---|---|---|---|---|---|
| SA | 5 | BA | 2 | DS | 1 |
| SB | 9 | BC | 1 | DC | 2 |
| SD | 6 | CS | 6 | DE | 2 |
| AB | 3 | CG2 | 5 | EG3 | 7 |
| AG1 | 9 | CF | 7 | FD | 2 |
| FG3 | 8 |
| Node | h(n) | Node | h(n) | Node | h(n) |
|---|---|---|---|---|---|
| S | 5 | D | 6 | G1 | 0 |
| A | 7 | E | 5 | G2 | 0 |
| B | 3 | F | 6 | G3 | 0 |
| C | 4 |
__________ is the heuristic function of greedy best-first search and __________ is heuristic function of A* Algorithmic search:
A) F(n)!=h(n) and f(n)=h(n)+g(n) B) F(n)=h(n) and f(n)=h(n)+g(n) C) F(n)>h(n) and f(n)=h(n)g(n) D) F(n)<h(n) and f(n)=h(n)+g(n)
The search strategy that uses problem-specific knowledge is known as:
A) Heuristic Search B) Informed Search C) Best-first Search D) All of the above
Apply A* algorithm on the following graph:
h(S)=15, h(A)=14, h(D)=12, h(B)=10, h(E)=10, h(C)=8, h(F)=10, h(G)=0
S is start node, G is goal node.
Graph: S→A(3), S→D(4), A→B(4), A→C(4), D→E(2), D→B(5), B→E(5), C→G(3), E→F(4), F→G(3)
3.2 · Local Search: Hill Climbing, Simulated Annealing, Genetic Algorithms
Explain Hill Climbing algorithm with an example. What problems occur in Hill Climbing? (P7 Q2C: What are the limitations of Hill Climbing Search and how can they be overcome?)
Explain the concept of Genetic Programming (Genetic Algorithms).
Write a short note on Simulated Annealing with a suitable example.
Write a short note on Genetic Algorithms.
What are local search algorithms? Explain any one in detail.
3.3 · Adversarial Search: Min-Max, Alpha-Beta Pruning
What do you understand by Min-Max Search and Alpha-Beta search? Explain in detail with example.
What is Game Playing Algorithm? Draw a game tree for the Tic-Tac-Toe problem.
What do you understand by α-β pruning? Apply alpha-beta pruning on the following game tree (root node = MAX):
Tree structure (appears across papers):
Root A (MAX)
├── B (MIN) → D: [2,3], E: [5,9]
└── C (MIN)
├── F: [0,1]
└── G: [7,5]
Leaf values: D=[2,3], E=[5,9], F=[0,1], G=[7,5]
Note: P4 paper has a larger tree with values: J–Z, a as leaf nodes with values 6,4,5,8,3,2,4,7,9,8,4,2,3,4,1,7,9,8
Write detailed note on Alpha Beta Pruning.
04
Knowledge and Reasoning
4.1 Knowledge-based Agents · Propositional Logic · First Order Logic (FOL): Syntax & Semantics · Inference in FOL · Forward Chaining · Backward Chaining
4.2 Knowledge Engineering in FOL · Unification · Resolution
4.3 Uncertain Knowledge & Reasoning · Uncertainty · Belief Networks · Simple Inference in Belief Network
4.2 Knowledge Engineering in FOL · Unification · Resolution
4.3 Uncertain Knowledge & Reasoning · Uncertainty · Belief Networks · Simple Inference in Belief Network
12 HRS
4.1 · First Order Logic & Quantifiers
Explain different quantifiers in FOL with example. (\(\forall x\), \(\exists x\), etc.)
What are you predicating by the logic \(\forall x : \exists y : \text{loyal\_to}(x,y)\)?
A) Everyone is loyal to all B) Everyone is loyal to someone C) Everyone is not loyal to someone D) Everyone is loyal
Convert the following into predicate / FOL form:
1. All vehicles have wheels. 2. Some-one speaks some language in class.
3. Everybody loves somebody sometimes. 4. All software engineers develop software.
5. Virat is a software engineer. (P7 also: All kings are persons; Every city in Maharashtra has temple; etc.)
3. Everybody loves somebody sometimes. 4. All software engineers develop software.
5. Virat is a software engineer. (P7 also: All kings are persons; Every city in Maharashtra has temple; etc.)
Compare and contrast Propositional Logic and First Order Logic.
Explain steps involved in converting a propositional logic statement into CNF with a suitable example.
Consider the following statements. Perform: i) Represent in FOL, ii) Convert each to CNF, iii) Prove using resolution technique (draw the resolution tree):
a) All people who are earning are happy.
b) All happy people smile.
c) Someone is earning. → Prove that someone is smiling.
b) All happy people smile.
c) Someone is earning. → Prove that someone is smiling.
Convert to FOL and CNF. Also prove using resolution:
"All people who are graduating are happy. All happy people smile. Someone is graduating. Prove that Is someone smiling?"
What actions would you take to prove "Some who are intelligent can't read" using propositional logic:
1. Whoever can read is literate. 2. Dolphins are not literate. 3. Some dolphins are intelligent.
The law says it is a crime for an American to sell weapons to hostile nations. The country Nono, an enemy of America, has missiles, and all of its missiles were sold to it by Colonel West, who is American. Prove that Col. West is a criminal using resolution.
Consider the facts: 1. Steve only likes easy courses. 2. Science courses are hard. 3. All courses in the basket-weaving department are easy. 4. BK301 is a basket-weaving course. Find by resolution that "What course would Steve like?"
4.1 · Forward & Backward Chaining
Illustrate forward-chaining and backward-chaining algorithm with a suitable example. Differentiate between the two.
4.2 · Resolution
What do you mean by Resolution? Also discuss the steps in Resolution.
Explain various methods of knowledge representation.
Which of the following is NOT a stage of knowledge engineering?
A) Assemble the relevant knowledge B) Encode general knowledge about the domain C) Identify the task D) Fixing a problem
4.3 · Uncertainty & Belief Networks
Define Belief Network. Describe the steps of constructing a belief network with an example.
Write a short note on conditional probability and its role in AI.
05
Planning and Learning
5.1 Planning: Planning Problem · State Space Search · Partial Order Planning · Hierarchical Planning · Conditional Planning
5.2 Learning: Forms & Theory of Learning · PAC Learning · Statistical Learning · Reinforcement Learning (Passive & Active)
5.2 Learning: Forms & Theory of Learning · PAC Learning · Statistical Learning · Reinforcement Learning (Passive & Active)
5 HRS
5.1 · Planning
What is planning in AI? Discuss some of the major approaches used in planning. Discuss partial order planning and hierarchical planning in detail.
Explain Partial-order planning with a suitable example.
Explain the concept of Conditional Order Planning.
Solve the Air Cargo Transport problem using STRIPS/Planning. It involves loading and unloading cargo onto/off planes and flying from place to place.
P2: Initial State: Cargo1 & Plane1 at SFO airport; Cargo2 & Plane2 at JFK airport. Goal: SFO has Cargo2, JFK has Cargo1.
P6: Initial State: Cargo1 & Plane1 at Mumbai; Cargo2 & Plane2 at Delhi. Goal: Cargo2 at Mumbai, Cargo1 at Delhi.
P6: Initial State: Cargo1 & Plane1 at Mumbai; Cargo2 & Plane2 at Delhi. Goal: Cargo2 at Mumbai, Cargo1 at Delhi.
Consider the problem of changing a flat tire. Goal: have a good spare tire mounted on axle. Initial state: flat tire on axle, good spare in trunk. Give the ADL description and discuss the solution.
What is present in an empty plan?
A) Start B) Finish C) Modest D) Both Start and Finish
Which is the most straightforward approach for planning?
A) Best first search B) Hill climbing search C) Depth first search D) State space search
5.2 · Learning — PAC, Reinforcement, Supervised
Explain the concept of PAC learning.
What do you understand by Reinforcement Learning? Explain in detail. (P2: Also explain the working of reinforcement learning.)
Write a short note on Reinforcement Learning.
Explain the concept of Supervised Learning.
06
AI Applications
A. NLP: Language Models · Grammars · Parsing
B. Robotics: Robots · Robot Hardware · Problems Robotics Can Solve
C. AI Applications in Healthcare, Retail, Banking & Wumpus World
4 HRS
A. Natural Language Processing
Explain different Language Models of Natural Language Processing.
Give types of parsing and generate the parse tree for the sentence "The cat ate the fish".
Explain the steps involved in Natural Language Processing.
B. Robotics
Illustrate / Describe application areas of AI in Robotics. How will AI help in Robotics applications?
C. AI Applications & Wumpus World
Write detailed notes on Applications of AI in Healthcare, Retail and Banking.
Write a detailed note on Wumpus World Environment. Explain giving its PEAS description and how the percept sequence is generated.