MCA-203 Artificial Intelligence and Machine Learning Notes cover all 4 units of this GGSIPU MCA Semester 3 paper in 67 pages, with 16 worked examples and numericals, 28 labelled diagrams and 76 exam-style questions. They follow the MCA syllabus for the batch admitted 2025-26 onwards, and the full notes come as a printed book.
Written by Prateek Talwar, Founder, Unnati Educations · Reviewed by Ritik Rawat, B.Tech, Senior Full Stack Developer · Checked against the MCA Scheme of Examination and Detailed Syllabus (A.S. 2025-26 onwards) on 13 October 2026 · Updated 14 October 2026 · Editorial policy
Scheme: MCA scheme and syllabus, A.S. 2025-26 onwards | Applies to: batch admitted 2025-26 onwards | Verified against: MCA Scheme of Examination and Detailed Syllabus (BoS 28/07/2025, AC Sub Committee 01/08/2025), 13 October 2026
MCA-203 Artificial Intelligence and Machine Learning at a glance: GGSIPU MCA Sem 3 credits, units and exam pattern
MCA-203 is one of two compulsory theory papers in MCA Semester 3, with MCA-201 Design and Analysis of Algorithms. Every figure below comes from the official MCA syllabus PDF on ipu.ac.in, so you can verify each one.
| Code and title | MCA-203 Artificial Intelligence and Machine Learning |
|---|---|
| Course and semester | MCA (2 years), Third Semester |
| Type and credits | Compulsory theory paper; L 3, T/P nil, 3 credits; 10 hours per unit |
| Marks | Out of 100; passing marks 40 out of 100. The internal share is not printed as a separate figure |
| Applies to | Batch admitted 2025-26 onwards |
| End-term pattern | Q1 compulsory, objective or short answers from the whole syllabus, 12 marks; then one of two questions from each of the 4 units, 12 marks each (12 + 48 = 60 marks) |
| Units | 4: AI Fundamentals; Knowledge Representation and Reasoning; Machine Learning; Unsupervised Learning |
| Paired lab | MCA-263 Artificial Intelligence and Machine Learning Lab.: L 0, P 2, 1 credit |
It sits with three electives and five labs among the GGSIPU MCA Semester 3 papers. The PDF names mathematical logic and linear algebra as prerequisites, which Unit 2 logic and Unit 4 PCA rely on.
The same PDF says the programme is governed by Ordinance 11, which holds the examination and result rules.
MCA-203 Artificial Intelligence and Machine Learning Notes unit-wise: 16 worked examples and numericals across 4 units
The official PDF opens each unit with a lead phrase, used here as the unit title. These MCA-203 AI and Machine Learning IPU notes keep the official unit order, so each H3 matches one unit question. Notes for every other MCA paper: GGSIPU MCA notes.
MCA-203 Unit 1: AI Fundamentals
Also covered as "AI Fundamentals and AI Search Techniques". Unit 1 (pages 4 to 22, 20 sections) covers AI definitions and Russell and Norvig's four approaches, AI types, the Turing test and the Chinese Room, AI vs ML vs deep learning, the water jug state space, then BFS, DFS, hill climbing, best-first, A*, AO*, CSP (map colouring, SEND + MORE = MONEY), means-end analysis and AI ethics. It has 5 worked examples, 8 diagrams, 12 short and 8 long questions.
| Step | Node expanded | What changes | OPEN after |
|---|---|---|---|
| 1 | S (g 0, f 6) | A at f 6, B at f 7 | A(6), B(7) |
| 2 | A (g 1, f 6) | B improves to 6; C at 8; G at 13 | B(6), C(8), G(13) |
| 3 | B (g 3, f 6) | C improves to 7 | C(7), G(13) |
| 4 | C (g 5, f 7) | G improves to 8, then G is reached | G(8) |
Quick look: A* expansion order with f = g + h for the cheapest S to G path (from page 16)
- Step
- 1
- Node expanded
- S (g 0, f 6)
- What changes
- A at f 6, B at f 7
- OPEN after
- A(6), B(7)
- Step
- 2
- Node expanded
- A (g 1, f 6)
- What changes
- B improves to 6; C at 8; G at 13
- OPEN after
- B(6), C(8), G(13)
- Step
- 3
- Node expanded
- B (g 3, f 6)
- What changes
- C improves to 7
- OPEN after
- C(7), G(13)
- Step
- 4
- Node expanded
- C (g 5, f 7)
- What changes
- G improves to 8, then G is reached
- OPEN after
- G(8)
Path S, A, B, C, G costs 8; with an admissible heuristic, A* returns this optimum.
MCA-203 Unit 2: Knowledge Representation and Reasoning
Unit 2 (pages 23 to 36, 15 sections) covers KR properties and types of knowledge, semantic networks and frames, propositional logic truth tables, FOPL translation of eight English sentences, inference rules with a truth-table proof of Modus Tollens, the seven CNF steps and the Marcus resolution proof, production rules, non-monotonic reasoning and uncertainty (Bayesian networks, certainty factors, fuzzy logic). It has 3 worked examples, 5 diagrams, 12 short and 8 long questions.
Quick look: Bayes' theorem on a rare-disease test (from page 33)
- Givens: P(D) = 0.01, P(+ | D) = 0.90, P(+ | not D) = 0.05.
- P(+) = 0.90 × 0.01 + 0.05 × 0.99 = 0.0585.
- P(D | +) = 0.009 / 0.0585 = 0.1538.
- A positive test means only about 15.4% chance of disease, because the prior is low.
MCA-203 Unit 3: Machine Learning
Also covered as "Machine Learning: Regression and Classification". Unit 3 (pages 37 to 51, 17 sections) covers Samuel and Mitchell definitions, ML types and applications, simple and multiple regression, least squares, TSS, SSR, SSE, R² and r on one five-point dataset, odds and the odds ratio, logistic regression, the confusion matrix and ROC, overfitting and the bias-variance trade-off. It has 4 solved numericals, 6 diagrams, 12 short and 6 long questions.
| Metric | Working | Value |
|---|---|---|
| Accuracy | (40 + 45) / 100 | 0.85 |
| Precision | 40 / (40 + 5) | 0.889 |
| Recall | 40 / (40 + 10) | 0.80 |
| F1-score | 2PR / (P + R) | 0.842 |
Quick look: confusion matrix metrics for 100 patients, TP 40, FN 10, FP 5, TN 45 (from page 47)
- Metric
- Accuracy
- Working
- (40 + 45) / 100
- Value
- 0.85
- Metric
- Precision
- Working
- 40 / (40 + 5)
- Value
- 0.889
- Metric
- Recall
- Working
- 40 / (40 + 10)
- Value
- 0.80
- Metric
- F1-score
- Working
- 2PR / (P + R)
- Value
- 0.842
MCA-203 Unit 4: Unsupervised Learning
Also covered as "Unsupervised Learning, Neural Networks and Recommender Systems". Unit 4 (pages 52 to 67, 15 sections) covers PCA on four points (PC1 keeps 92.1% of variance), K-means with k = 3, bagging, Random Forest and AdaBoost, the perceptron and NN types, seven activation functions, a 3-4-2 feed-forward net, one back-propagation iteration and content-based recommendation with TF-IDF and cosine similarity. It has 4 solved numericals, 9 diagrams, 12 short and 6 long questions.
| Movie | Genre vector | Cosine similarity |
|---|---|---|
| Movie B | (1, 0, 1, 0) | 0.500 |
| Movie C | (0, 1, 0, 1) | 0.500 |
| Movie D | (1, 1, 1, 0) | 0.816, recommended first |
Quick look: which movie a content-based recommender picks, user profile (1, 1, 0, 0) (from page 64)
- Movie
- Movie B
- Genre vector
- (1, 0, 1, 0)
- Cosine similarity
- 0.500
- Movie
- Movie C
- Genre vector
- (0, 1, 0, 1)
- Cosine similarity
- 0.500
- Movie
- Movie D
- Genre vector
- (1, 1, 1, 0)
- Cosine similarity
- 0.816, recommended first
The "View Sample PDF" button above opens one content page from each unit, plus the cover and index.
Forward and backward reasoning in AI: excerpt from page 31 of the MCA-203 notes
Excerpt from page 31
Types of reasoning
Reasoning is the process of drawing conclusions and making predictions from the knowledge, facts and beliefs a system already holds. Deductive reasoning moves from a general rule to a specific conclusion and is always correct when the premises are true: all men are mortal, Socrates is a man, so Socrates is mortal.
Inductive reasoning moves from specific observations to a general rule, so its result is probable rather than certain: every crow seen so far is black, so all crows are black. Abductive reasoning picks the most likely explanation for an observation (the road is wet, so it probably rained). Common-sense reasoning uses everyday experience, such as knowing that touching fire burns.
Monotonic reasoning never withdraws a conclusion when new facts arrive, which suits theorem proving. Non-monotonic reasoning may withdraw one: Tweety flies until we learn Tweety is a penguin.
Forward reasoning (forward chaining)
Forward reasoning is data-driven and works bottom-up. The inference engine starts from the known facts, fires every rule whose IF-part is satisfied, adds that rule's THEN-part to working memory as a new fact, and keeps going until the goal appears or no rule can fire.
Backward reasoning (backward chaining)
Backward reasoning is goal-driven and works top-down. It starts from the hypothesis to be proved, looks for rules whose THEN-part matches it, and turns their IF-parts into sub-goals. It continues until every sub-goal is matched with a known fact.
Solved example
Rules: R1: A ∧ B → C; R2: C → D; R3: D ∧ E → F. Facts: A, B, E. Goal: prove F.
Forward chaining: facts A and B satisfy R1, so C is added. C satisfies R2, so D is added. D and the fact E satisfy R3, so F is added and the goal is reached. Figure 2.3 in the notes draws this chain from data to goal, and the next page runs the same rules backwards from F.
MCA-203 exam corner: 48 short-answer and 28 long-answer questions with 16 worked examples by unit
Each unit ends with a terms or formula table, a likely-question list and a quick revision table. The questions are practice prompts; their answers come from the sections they point to.
| Unit | Short-answer (notes label: 2.5 marks) | Long-answer (notes label: 12.5 marks) | Worked examples |
|---|---|---|---|
| 1 AI Fundamentals | 12 | 8 | 5 |
| 2 Knowledge Representation and Reasoning | 12 | 8 | 3 |
| 3 Machine Learning | 12 | 6 | 4 |
| 4 Unsupervised Learning | 12 | 6 | 4 |
| Total | 48 | 28 | 16 |
MCA-203 exam-style questions and worked examples per unit
- Unit
- 1 AI Fundamentals
- Short-answer (notes label: 2.5 marks)
- 12
- Long-answer (notes label: 12.5 marks)
- 8
- Worked examples
- 5
- Unit
- 2 Knowledge Representation and Reasoning
- Short-answer (notes label: 2.5 marks)
- 12
- Long-answer (notes label: 12.5 marks)
- 8
- Worked examples
- 3
- Unit
- 3 Machine Learning
- Short-answer (notes label: 2.5 marks)
- 12
- Long-answer (notes label: 12.5 marks)
- 6
- Worked examples
- 4
- Unit
- 4 Unsupervised Learning
- Short-answer (notes label: 2.5 marks)
- 12
- Long-answer (notes label: 12.5 marks)
- 6
- Worked examples
- 4
- Unit
- Total
- Short-answer (notes label: 2.5 marks)
- 48
- Long-answer (notes label: 12.5 marks)
- 28
- Worked examples
- 16
The 2.5 and 12.5 labels do not match the 2025-26 paper-setter instructions, where Q1 carries 12 marks in total and each unit question carries 12 marks. Plan answers to the official pattern.
The MCA-203 notes list 48 short-answer and 28 long-answer exam-style questions across 4 units, plus 16 worked examples and numericals.
One question per unit, with answer pointers:
- Unit 1: Explain AO* for AND-OR graphs. How is it different from A*? Pointer (page 17): AO* returns a solution sub-graph and revises costs backwards (minimum over OR, sum over AND arcs); A* returns one path.
- Unit 2: Prove that Modus Tollens is valid using a truth table. Pointer (page 28): the column [(P → Q) ∧ ¬Q] → ¬P is true in all four rows, so it is a tautology.
- Unit 3: Define odds and odds ratio. Pointer (pages 44 to 45): odds = p / (1 − p); from a 2 × 2 table the odds ratio is ad / bc, which gives 3.86 in the smokers example.
- Unit 4: Why can a single perceptron not learn XOR? Pointer (page 59): a perceptron draws only a linear boundary and XOR is not linearly separable, which led to multi-layer networks.
More practice sits in the MCA-203 important questions.
Who writes and checks the MCA-203 AI and ML notes?
- Prateek Talwar, author: Founder, Unnati Educations; Associate Company Secretary (ACS), ICSI, membership no. 58704, since 15 May 2019; B.Com, University of Delhi (School of Open Learning), 2012. Writes and edits the MCA-203 notes. Author profile
- Ritik Rawat, subject reviewer: Senior Full Stack Developer, Unnati Educations; B.Tech. Reviews the search traces, ML numericals and back-propagation working.
The 2026-27 edition of these GGSIPU MCA Sem 3 Artificial Intelligence and Machine Learning notes follows the MCA scheme approved on 01 August 2025, with Russell and Norvig (TB1) as the main reference. Last updated 14 October 2026; see the editorial policy.
The same team prepares GGSIPU important questions for every course, so the MCA-203 question bank keeps this unit order.
Do the MCA-203 notes cover the full GGSIPU syllabus?
Yes. Every topic the official PDF lists for the four MCA-203 units has its own section in the notes; the unit check found no gaps.
| Official unit | Official topics | Where in the notes | Status |
|---|---|---|---|
| Unit I: AI Fundamentals | Defining AI, types of AI, Turing test, AI techniques, AI vs ML vs deep learning, problem characterization and reduction, state space, hill climbing, BFS, DFS, best-first, A*, AO*, constraint satisfaction, MEA, ethics of AI | Sections 1.1 to 1.17, pages 5 to 20 | Covered |
| Unit II: Knowledge Representation and Reasoning | Approaches to KR, propositional logic, FOPL, inference rules (Modus Ponens, Modus Tollens, resolution, And elimination, syllogism), production rules, types of knowledge, forward and backward reasoning, non-monotonic reasoning, reasoning with uncertainties | Sections 2.1 to 2.12, pages 24 to 34 | Covered |
| Unit III: Machine Learning | Introduction, why ML, types of ML problems, applications; simple and multiple regression, least squares, TSS, SSR, SSE, odds, odds ratio; logistic regression; coefficient of determination, correlation, confusion matrix; overfitting, underfitting, bias and variance | Sections 3.1 to 3.14, pages 38 to 49 | Covered |
| Unit IV: Unsupervised Learning | Dimensionality reduction (PCA), K-means, ensemble learning (boosting and bagging), neural networks and their types, activation functions, feed forward, back propagation, recommender systems, content-based recommendations | Sections 4.1 to 4.12, pages 53 to 65 | Covered |
MCA-203 official topics mapped to the notes
- Official unit
- Unit I: AI Fundamentals
- Official topics
- Defining AI, types of AI, Turing test, AI techniques, AI vs ML vs deep learning, problem characterization and reduction, state space, hill climbing, BFS, DFS, best-first, A*, AO*, constraint satisfaction, MEA, ethics of AI
- Where in the notes
- Sections 1.1 to 1.17, pages 5 to 20
- Status
- Covered
- Official unit
- Unit II: Knowledge Representation and Reasoning
- Official topics
- Approaches to KR, propositional logic, FOPL, inference rules (Modus Ponens, Modus Tollens, resolution, And elimination, syllogism), production rules, types of knowledge, forward and backward reasoning, non-monotonic reasoning, reasoning with uncertainties
- Where in the notes
- Sections 2.1 to 2.12, pages 24 to 34
- Status
- Covered
- Official unit
- Unit III: Machine Learning
- Official topics
- Introduction, why ML, types of ML problems, applications; simple and multiple regression, least squares, TSS, SSR, SSE, odds, odds ratio; logistic regression; coefficient of determination, correlation, confusion matrix; overfitting, underfitting, bias and variance
- Where in the notes
- Sections 3.1 to 3.14, pages 38 to 49
- Status
- Covered
- Official unit
- Unit IV: Unsupervised Learning
- Official topics
- Dimensionality reduction (PCA), K-means, ensemble learning (boosting and bagging), neural networks and their types, activation functions, feed forward, back propagation, recommender systems, content-based recommendations
- Where in the notes
- Sections 4.1 to 4.12, pages 53 to 65
- Status
- Covered
Official source: MCA scheme of examination and detailed syllabus (A.S. 2025-26 onwards).
The PDF lists R² and correlation under classification; the notes teach them in section 3.8 on the regression dataset (R² = 0.60, r = 0.775). Every course in GGSIPU notes is mapped to its official PDF the same way.
How to write MCA-203 answers: algorithm traces, step-wise numericals and twelve-mark theory answers
MCA-203 asks for three kinds of answer. Q1 wants short, exact points from all four units. The unit questions want an algorithm with a trace (A*, AO*, resolution, K-means, back-propagation), a numerical (least squares, odds ratio, PCA, confusion matrix), or both as subparts adding up to 12 marks.
Algorithm answers
The Unit 1 exam tip sets the order: definition, algorithm or steps, diagram or worked example, advantages and disadvantages, then a comparison such as BFS vs DFS. For resolution, the Unit 2 tip splits the work into FOPL sentences, clause form, the negated goal and a resolution tree that ends in the empty clause.
Numerical answers
Show the working table before the result.
Worked example: least squares line for (1, 2), (2, 4), (3, 5), (4, 4), (5, 5) (page 42)
- Means: x̄ = 15 / 5 = 3 and ȳ = 20 / 5 = 4.
- Table columns give Σ(x − x̄)(y − ȳ) = 6 and Σ(x − x̄)² = 10.
- Slope b1 = 6 / 10 = 0.6; intercept b0 = 4 − 0.6 × 3 = 2.2.
- Line ŷ = 2.2 + 0.6x, so x = 6 predicts 5.8.
For K-means, the Unit 4 tip asks for a full distance table in iteration 1, a summary table for later iterations, and the word "converged" once clusters stop changing.
Slips the notes flag
- Writing ∀x Student(x) ∧ Passes(x) instead of pairing ∀ with →.
- Trusting accuracy alone on imbalanced fraud data.
- Reading a high r as proof that x causes y.
Practise this format on GGSIPU MCA important questions.
MCA-203 and the MCA-263 AI and ML lab (1 credit), plus other GGSIPU MCA Sem 3 papers
Paired lab: MCA-263 Artificial Intelligence and Machine Learning Lab., L 0, P 2, 1 credit. The official PDF gives its six course outcomes but no fixed practical list. How they map to theory units:
- CO1 heuristic search on puzzles: Unit 1, sections 1.8 to 1.15. CO5 inference rules for a knowledge base: Unit 2, sections 2.6 to 2.8.
- CO3 bias and regression, CO4 evaluating classifiers: Unit 3, sections 3.5 to 3.14.
- CO2 dimensionality reduction: Unit 4, section 4.3 (PCA). CO6 a domain-specific intelligent application can build on the recommenders in 4.11 to 4.12.
Other Semester 3 papers: MCA-201 Design and Analysis of Algorithms (compulsory). Core Elective II, one of: MCA-205 Statistics and Data Analytics, MCA-207 Enterprise Computing with JAVA, MCA-209 Natural Language Processing, MCA-211 Computer Graphics, MCA-213 Wireless Sensor Networks, MCA-215 Software Project Management, MCA-217 Advanced Computer Architecture, MCA-219 Distributed Systems, MCA-221 Applied Cryptography, MCA-223 Cloud Computing, MCA-225 e-Business Systems.
Core Elective III, one of: MCA-227 Web Intelligence and Big Data, MCA-229 Flutter and Dart, MCA-231 Service Oriented Architecture, MCA-233 Multimedia Technologies, MCA-235 Internet of Things, MCA-237 Soft Computing, MCA-239 Software Quality Management, MCA-241 Digital Image Processing, MCA-243 Compiler Design, MCA-245 Parallel Computing, MCA-247 Numerical and Scientific Computing.
Open Elective II, one of: MCA-249 Research Methodology, MCA-251 Operational Research, MCA-253 Cyber Security and Cyber Laws, MCA-255 e-Content Development. Labs MCA-261 to MCA-267, Minor Project III (MCA-269) and NUES courses MCA-271 and MCA-273 complete the term. MCA-209 and MCA-237 reuse the Unit 4 neural network pages.
MCA-203 AI and ML notes FAQs
How do I get the full MCA-203 notes?
Message us on WhatsApp with the code MCA-203 using the "Get Full Notes" button. The full notes come only as a printed book, not as a PDF: 67 pages covering all 4 units, with 16 worked examples and numericals. Delivery takes 5 to 7 days at most; the exact time depends on your location. Check the sample PDF above first.
Are the MCA-203 notes the same as notes for the MCA-263 lab?
No. MCA-203 is the 3-credit theory paper with a written end-term exam; MCA-263 is the 1-credit lab (L 0, P 2) where the same ideas are implemented. These MCA-203 notes teach the theory, traces and numericals. The lab has its own notes page, and its course outcomes line up with units here, for example PCA in Unit 4.
Which MCA-203 unit has the most worked examples?
Unit 1 has the most: 5 worked examples (water jug, A*, AO*, map colouring and SEND + MORE = MONEY) across 19 pages. Units 3 and 4 each carry 4 solved numericals, from least squares and odds ratio to PCA, K-means, back-propagation and cosine similarity. Unit 2 has 3, including the Marcus resolution proof and a Bayes' theorem problem.
Do the MCA-203 notes follow the 2025-26 MCA syllabus?
Yes. They were checked on 13 October 2026 against the MCA scheme of examination and detailed syllabus effective from Academic Year 2025-26, approved by the BoS on 28 July 2025. All four units and every listed topic, from the Turing test to content-based recommendations, have their own section. They suit the batch admitted 2025-26 onwards.
What is the passing mark for MCA-203?
MCA-203 is marked out of 100, and its end-term question pattern totals 60 marks: 12 for Q1 plus 12 for each of four unit questions. The MCA scheme of examination states pass marks as 40%, which is 40 out of 100 for every paper. For end-term minimums, re-appear and result rules, see the university's Ordinance 11 on ipu.ac.in.
Why do the MCA-203 notes label questions 2.5 and 12.5 marks?
Those labels sit beside the likely-question lists and do not match the 2025-26 MCA-203 paper-setter instructions, which set Q1 at 12 marks overall and each unit question at 12 marks. Treat the lists as topic practice and size your answers to the official 12-mark pattern, splitting long answers into subparts where a question has them.
Get the full MCA-203 Artificial Intelligence and Machine Learning Notes on WhatsApp
| Unit | Pages | Sections | Worked examples | Diagrams |
|---|---|---|---|---|
| 1 AI Fundamentals | 4 to 22 (19) | 20 | 5 | 8 |
| 2 Knowledge Representation and Reasoning | 23 to 36 (14) | 15 | 3 | 5 |
| 3 Machine Learning | 37 to 51 (15) | 17 | 4 | 6 |
| 4 Unsupervised Learning | 52 to 67 (16) | 15 | 4 | 9 |
What the full MCA-203 notes book contains, unit by unit
- Pages
- 4 to 22 (19)
- Sections
- 20
- Worked examples
- 5
- Diagrams
- 8
- Pages
- 23 to 36 (14)
- Sections
- 15
- Worked examples
- 3
- Diagrams
- 5
- Pages
- 37 to 51 (15)
- Sections
- 17
- Worked examples
- 4
- Diagrams
- 6
- Pages
- 52 to 67 (16)
- Sections
- 15
- Worked examples
- 4
- Diagrams
- 9
These Artificial Intelligence and Machine Learning notes in English run to 67 pages with a page-numbered index, 11 quotable definitions and a quick revision table closing every unit. They are available only as a printed book, not as a PDF. Delivery: within 5 to 7 days at most; the exact time may vary with your location.
Get the full MCA-203 notes book on WhatsApp (4 units, 67 pages)
Unnati Educations is an independent education provider based in Vikaspuri, Delhi. We are not affiliated with, authorised by, or endorsed by Guru Gobind Singh Indraprastha University (GGSIPU). 'GGSIPU' and 'IPU' are used only to describe the syllabus our notes follow. For official information, visit ipu.ac.in.
Ready for the full notes?
Discuss coverage and delivery with our team.