Previous-year questions, by module
Every question from the six Mumbai University papers (May 2023 to May 2026), sorted by syllabus topic. Repeats are merged into one entry; other wordings sit under “Also asked as”.
How the paper works
- 3 hours, 80 marks. Same format in all six papers.
- Q1 is compulsory: answer any 4 of 5 short questions.
- Q2 to Q6: answer any 3. Each has two parts (a, b).
- Q6 is usually “detailed note on any two” of three topics.
- Each paper has one or two numericals with a table.
Times asked, by module
Asked in 4 or more of the 6 papers
- 6×M5 · 5.4Explain the DBSCAN clustering algorithm along with an example.
- 5×M2 · 2.3Explain logistic regression.
- 5×M3 · 3.7Explain the Random Forest algorithm in detail.
- 5×M6 · 6.3Explain Linear Discriminant Analysis (LDA) for dimension reduction, with an example.
- 4×M1 · 1.4Explain any five applications of Machine Learning.
- 4×M2 · 2.1Explain regression line, scatter plot, error in prediction and best fitting line.
Introduction to Machine Learning
1.2Types of Machine Learning
-
1×Differentiate between supervised and unsupervised learning.
-
1×Differentiate between regression and classification, with examples.
1.3Issues in Machine Learning
-
3×Explain the issues in Machine Learning.Also asked as
- Write a short note on issues in Machine Learning. May 24
- What are the issues in Machine Learning? May 25
1.4Applications of Machine Learning
-
4×Explain any five applications of Machine Learning.Also asked as
- Explain any five business applications of Machine Learning. May 24
1.5Steps in developing a Machine Learning application
-
3×Explain the steps involved in developing a Machine Learning application.Also asked as
- Demonstrate steps to design a Machine Learning application. Dec 24
- Explain the stages involved in developing a Machine Learning system. May 26
-
1×How do you choose the right ML algorithm?
1.6Training error, generalization error
-
1×Explain training error and generalization error.
1.7 · 1.8Overfitting, underfitting, bias-variance trade-off
-
1×What is overfitting, underfitting and the bias-variance trade-off in Machine Learning?
-
1×Explain the bias-variance trade-off with a suitable example.
1.1 basics was never asked on its own.
Learning with Regression and Trees
2.1Linear regression
-
4×Explain regression line, scatter plot, error in prediction and best fitting line.
-
1×Write a detailed note on linear regression.
-
2×List and explain the performance evaluation measures used in regression problems.Also asked as
- List and explain any four performance metrics used in regression problems. May 26
- Asked as the second half of the mass vs time numerical below. May 24
-
1×The mass y (grams) of a chemical is related to the time x (seconds) for which the reaction has been taking place, as in the table. Find the equation of the regression line. Also explain performance evaluation measures for regression.
Time, x (seconds) 5 7 12 16 20 Mass, y (grams) 40 120 180 210 240 -
1×The table shows midterm and final exam grades of students in a database course. Using linear regression, predict the final exam grade of a student who got 86 in the midterm.
Midterm exam (X) 72 50 81 74 94 86 59 83 86 33 88 81* Final exam (Y) 84 53 77 78 90 75 49 79 77 52 74 90 * The last midterm value is faint in the scan. It reads as 81, which matches the textbook version of this dataset (Han & Kamber, Data Mining). Check against your copy.
-
1×Explain simple linear regression. Find the regression line Y = aX + C using the least squares method and estimate the expenditure of the company in the 6th month. Monthly expenditure (in thousands):
X (Month) 1 2 3 4 5 Y (Expenditure) 12 19 29 37 45
2.2Multivariate linear regression
-
1×Explain the multivariate linear regression method.
2.3Logistic regression
-
5×Explain logistic regression.Also asked as
- Explain the concept of logistic regression. May 23
- Explain logistic regression and performance evaluation metrics. Dec 24
- Write a detailed note on logistic regression. Dec 25, May 26
-
1×Differentiate between linear regression and logistic regression.
-
2×Consider the use case of email spam detection. Identify and explain the suitable machine learning technique for this task.
No single syllabus topic. Spam vs not-spam is binary classification, so logistic regression (2.3) is the most direct fit; SVM (4.5) also works.
2.4 · 2.5Decision trees and the Gini index
-
2×Describe decision trees. Explain how trees are constructed using the Gini index, with an example.Also asked as
- Explain the Gini index along with an example. May 24
-
1×Explain the concepts of entropy and Gini.
-
1×Create a decision tree using the Gini index to classify the following dataset for profit.
Age Competition Type Profit old Yes software Down old No software Down old No hardware Down mid Yes software Down mid Yes hardware Down mid No hardware Up mid No software Up new Yes software Up new No hardware Up new No software Up -
1×The dataset below has 3 features (Color, Wig, Num. Ears) and one output variable, Emotion.
- Find the root node of the decision tree using the Gini index.
- Explain techniques that can be used to handle overfitting in decision trees.
Color G G G B B R R R R Wig Y N N N N N N N Y Num. Ears 2 2 2 2 2 2 2 2 3 Emotion S S S S H H H H H -
1×Use the Gini index to decide whether a car will be stolen. The target is “car is stolen?” (Yes or No). Create a decision tree for the data below.
Car no Colour Type Origin Stolen? 1 Red Sports Domestic Yes 2 Red Sports Domestic No 3 Red Sports Domestic Yes 4 Yellow Sports Domestic No 5 Yellow Sports Imported Yes 6 Yellow SUV Imported No 7 Yellow SUV Imported Yes 8 Yellow SUV Domestic No 9 Red SUV Imported No 10 Red Sports Imported Yes
2.6CART
-
1×Demonstrate the CART method along with an example.
2.7 – 2.11Performance metrics
-
3×Explain performance evaluation metrics for classification, with a suitable example.Also asked as
- Explain performance evaluation metrics for binary classification with a suitable example. May 24
- Write a detailed note on performance metrics for classification. May 25
-
1×Explain the following performance metrics with an example: accuracy, precision, recall and F1 score.
-
1×You are analysing data on 50 people applying for a grant. Each proposal was read by two readers, and each said “Yes” or “No”. Calculate precision, recall and accuracy.
Reader B (Predicted) Yes No Reader A (Actual) Yes 20 5 No 10 15 -
1×Explain the confusion matrix. Calculate accuracy, precision, recall and F1 score for the following:
Predicted + Predicted − Actual + 40 8 Actual − 12 30
2.8 Kappa statistics and 2.11 ROC curve were never asked by name.
Ensemble Learning
3.2K-fold cross validation
-
2×Explain the concept of k-fold cross validation.Also asked as
- Asked together with XGBoost: “Write a short note on XGBoost and explain k-fold cross validation.” May 26
3.3 · 3.6 · 3.8Bagging and boosting
-
2×Explain the concept of bagging and boosting.
-
1×Compare bagging and boosting with reference to ensemble learning. Explain how these methods help improve the performance of a machine learning model.
3.5XGBoost
-
3×Write a short note on the XGBoost ensemble method.
3.7Random forest
-
5×Explain the Random Forest algorithm in detail.Also asked as
- Demonstrate the ensemble-learning-based Random Forest algorithm in detail. Dec 24
- Explain the ensemble learning algorithm Random Forest and its use cases in real-world applications. May 24
3.9Ways to combine classifiers
-
1×Explain the different ways to combine classifiers.
3.1 ensembles (alone), 3.4 stumping and subagging were never asked.
Learning with Classification
4.2Optimal decision boundary, margins and support vectors
-
1×Define the following terms with reference to SVM: hyperplane, support vectors, hard margin, soft margin, kernel.
-
1×Explain the concept of margin and support vector.
-
1×What is the significance of the optimal separating hyperplane in SVM?
4.1 · 4.3SVM as a constrained optimization problem
-
1×Explain the support vector machine as a constrained optimization problem.
4.6Kernel trick
-
3×Explain the SVM kernel trick.Also asked as
- Explain kernel functions of SVM along with an example. Dec 25
- Write a short note on support vector machine and the SVM kernel trick. May 26
4.8Multiclass classification
-
3×Describe multiclass classification.
4.4 quadratic programming, 4.5 linear vs nonlinear SVM and 4.7 support vector regression were never asked.
Learning with Clustering
5.1Introduction to clustering: distance metrics, major approaches
-
1×Explain the distance metrics used in clustering.
-
1×Explain the K-means algorithm.
K-means is not named in the syllabus. It fits under 5.1 as the standard partitioning approach.
5.2Graph-based clustering: minimal spanning tree
-
3×Explain clustering with a minimal spanning tree (MST), with an example.Also asked as
- Demonstrate the MST algorithm along with an example. Dec 24, May 25
5.3Model-based clustering: Expectation Maximization
-
3×Explain the concept of the Expectation Maximization algorithm.
5.4Density-based clustering: DBSCAN
-
6×Explain the DBSCAN clustering algorithm along with an example.Also asked as
- What is density-based clustering? Explain the steps used for clustering with DBSCAN. May 24
- Write a detailed note on DBSCAN. May 23, May 25
-
1×Describe the DBSCAN algorithm. Apply DBSCAN on the dataset {1, 2, 3.5, 4, 5, 11} with ε = 2 and MinPts = 4.
Dimensionality Reduction
6.1Dimensionality reduction techniques
-
1×What is dimensionality reduction? Explain how it can be used for classification and clustering tasks.
-
1×Explain the concept of feature selection and feature extraction.
6.2Principal Component Analysis (PCA)
-
3×Describe the PCA algorithm along with an example.Also asked as
- Give the steps to design the PCA dimensionality reduction technique along with an example. Dec 24
- Write a detailed note on PCA for dimension reduction. May 25
6.3Linear Discriminant Analysis (LDA)
-
5×Explain Linear Discriminant Analysis (LDA) for dimension reduction, with an example.Also asked as
- Explain LDA and its real-world applications. May 24
- Write a short note on linear discriminant projection along with an example. May 25
- Write a detailed note on LDA for dimension reduction. May 23, May 26
6.4Singular Value Decomposition (SVD)
-
1×Find the SVD of the matrix below.A =22−11