Skip to content
EndsemNov 20252019 patternTE Comp

Machine Learning (ML) SPPU Endsem Question Paper Nov 2025

All 8 questions from this SPPU endsem paper, as printed. Read them here or open the PDF.

Open the PDFMost repeated ML questions
Subject code
410242
Pattern
2019 pattern
Exam
Endsem
Session
Nov 2025
Max. marks
70
Time
2½ Hours
Semester
VII
Paper code
[6584]-92

Printed for: B.E. (Computer Engineering)

Instructions to the candidates

  1. Answer Q1 or Q2, Q3 or Q4, Q5 or Q6, Q7 or Q8.
  2. Figures to the right side indicate full marks.
  3. Draw neat diagram wherever necessary.
  4. Assume suitable data, if necessary.

Questions

  1. Q1

    1. a)

      What is R2 measure of evaluation?

      Also asked:May 2024

      [6]
    2. b)

      Explain the concept of a cost function and its role in Gradient Descent.

      [6]
    3. c)

      What is the primary purpose of the L2 regularization term in the Ridge Regression cost function?

      [6]

  2. Q2

    1. a)

      Describe the bias-variance trade-off and its relationship to underfitting and overfitting.

      [6]
    2. b)

      Month Amount spent Sales Mar-23 13447.45 5750.00 Feb-23 10692.38 44305.00 Jan-23 3067.05 34980.00 Dec-22 200.00 10500.00 Nov-22 12343.18 39154.00 Oct-22 12300.18 23065.00 Sep-22 11960.35 20745.00 Aug-22 9924.53 42000.00 Jul-22 8873.12 5750.00 Jun-22 6383.40 41500.00 May-22 1397.60 13500.00 The table shows advertisement spent and sales figure for 11 months. Predict what would be sales if ad spend is 10800/- using linear regression.

      [6]
    3. c)

      What are different techniques to reduce overfitting? 2

      [6]

  3. Q3

    1. a)

      Explain the basic principles of a Support Vector Machine (SVM). What are support vectors and how does SVM find the optimal hyper plane for classification?

      [5]
    2. b)

      What is Multi Class Classification? Explain the variants of Multi Class Classification.

      Also asked:May 2023

      [6]
    3. c)

      You have a binary classification model with the following confusion matrix: True Positives (TP): 120 True Negatives (TN): 45 False Positives (FP): 10 False Negatives (FN): 5 Calculate the accuracy, precision, recall, and F1-score of the model.

      [6]

  4. Q4

    1. a)

      With suitable diagram, Explain Random forest Algorithm with example.

      [5]
    2. b)

      What are different distance metrics are used in K-NN?

      Also asked:May 2024

      [6]
    3. c)

      Why ensemble learning is used for ML?

      Also asked:May 2025 ·May 2024

      [6]

  5. Q5

    1. a)

      What is K-means clustering, and how does it work? Explain the basic principles of this unsupervised machine learning algorithm and its applications.

      [5]
    2. b)

      Why K-medoid is used? Explain K-medoid algorithm.

      [6]
    3. c)

      Define the terms Core point, Border point, Noise point, Eps (  ), MinPts (min - samples), Reachability, Density-connected used in the Density- Based Spatial Clustering (DBSCAN) algorithm.

      [6]

  6. Q6

    1. a)

      What is isolation factor model?

      Also asked:May 2024

      [5]
    2. b)

      Write short note on following Hierarchical clustering method : i) Agglomerative ii) Dendogram

      [6]
    3. c)

      Explain the intrinsic and extrinsic methods for measuring clustering quality. 3

      [6]

  7. Q7

    1. a)

      Describe the architecture of Artificial Neural Network.

      [6]
    2. b)

      Explain the Recurrent Neural Network (RNN) with suitable example.

      Also asked:Nov 2023

      [6]
    3. c)

      Explain building blocks of RBF networks.

      Also asked:May 2024 ·May 2023

      [6]

  8. Q8

    1. a)

      What is multilayer perceptron? Describe with diagram.

      [6]
    2. b)

      Explain the Convolution Neural Network (CNN) with suitable example.

      Also asked:May 2024

      [6]
    3. c)

      What is the role of activation functions in neural networks, and why are they necessary for the network’s learning process? ♦♦♦

      [6]