AL-702-D Machine Learning for Data Science - Jun 2025 Question Paper

Under each question, How to answer lists the points an answer should cover, outlined by CampusPrep from the syllabus. Check the details against your textbook.

1. 7 Marks
   aDescribe the main types of machine learning algorithms and their applications in data science.
   How to answer

   • Define supervised, unsupervised, semi-supervised and reinforcement learning
   • Give example algorithms for each type e.g. regression, clustering, Q-learning
   • Describe data science applications e.g. prediction, segmentation, recommendation
   • Summarize selection criteria based on data and task
2. 7 Marks
   bExplain how the randomization technique can improve algorithm performance.
   How to answer

   • Define randomization technique in algorithms
   • Explain avoidance of worst-case and local optima
   • Discuss improved generalization and robustness
   • Give examples like random initialization, dropout, random forests
3. 7 Marks
   aWhat is the role of hash tables in search algorithms?
   How to answer

   • Define hash table and hash function
   • Explain O(1) average search, indexing and lookup
   • Discuss collision handling and use in large datasets
   • Give search algorithm examples
4. 7 Marks
   bDiscuss the applications of dynamic programming in personal genomics.
   How to answer

   • Define dynamic programming principles
   • Explain sequence alignment and variant analysis
   • Discuss optimal substructure in genomics problems
   • Give examples like Needleman-Wunsch, Smith-Waterman
5. 7 Marks
   aExplain the concept of NP completeness in the context of algorithms.
   How to answer

   • Define P, NP and NP-complete classes
   • Explain polynomial-time verification and reduction
   • Give examples like SAT, TSP, knapsack
   • Discuss relevance to algorithm design and intractability
6. 7 Marks
   bDescribe the significance of data science in analyzing massive genomic data.
   How to answer

   • Describe volume and complexity of genomic data
   • Explain ML and statistical methods for pattern discovery
   • Discuss variant calling, disease prediction, personalization
   • Highlight scalability and visualization role
7. 7 Marks
   aWhat is ensemble learning? Explain its advantages in predictive modeling.
   How to answer

   • Define ensemble learning and combining learners
   • Explain bagging, boosting and stacking
   • Discuss reduction in variance, bias and overfitting
   • Give advantages in accuracy and robustness for prediction
8. 7 Marks
   bDescribe the steps involved in cross-validation and its role in model accuracy.
   How to answer

   • Define cross-validation purpose
   • List steps: split into k-folds, train, validate, average
   • Explain types like k-fold, leave-one-out
   • Discuss role in estimating accuracy and preventing overfitting
9. 7 Marks
   aDefine linear classification and probabilistic modeling in machine learning.
   How to answer

   • Define linear classification with decision boundary
   • Define probabilistic modeling with probability distributions
   • Give examples like logistic regression, Naive Bayes
   • Contrast deterministic vs probabilistic outputs
10. 7 Marks
    bExplain how model selection is essential for machine learning performance.
    How to answer

    • Define model selection problem
    • Explain bias-variance tradeoff and generalization
    • Discuss validation, metrics and hyperparameter tuning
    • Link to improved performance and avoiding under/overfitting
11. 7 Marks
    aWhat are topic models and how are they applied in data analysis?
    How to answer

    • Define topic models for discovering latent topics
    • Explain LDA and probabilistic topic assignment
    • Describe applications in text mining and document clustering
    • Give data analysis workflow example
12. 7 Marks
    bDiscuss the concept of probabilistic inference in data science.
    How to answer

    • Define probabilistic inference and Bayes rule
    • Explain estimating posteriors from data
    • Discuss handling uncertainty in predictions
    • Give data science examples
13. 7 Marks
    aExplain the importance of data cleaning in data preprocessing.
    How to answer

    • Define data cleaning in preprocessing
    • Explain handling missing, noisy and inconsistent data
    • Discuss impact on model quality and reliability
    • List techniques like deduplication and outlier treatment
14. 7 Marks
    bDiscuss how data transformation affects model building in machine learning.
    How to answer

    • Define data transformation types
    • Explain normalization, encoding and feature scaling
    • Discuss effect on convergence and model accuracy
    • Give examples of improved model building
15. 7 Marks
    aDefine OLAP and OLTP and their applications in data warehousing.
    How to answer

    • Define OLAP for analytical processing and OLTP for transactions
    • Contrast characteristics: query type, speed, schema
    • Describe data warehousing applications like reporting and mining
    • Give examples of use cases
16. 7 Marks
    bExplain the connection between data warehousing and data mining in data science.
    How to answer

    • Define data warehousing and data mining
    • Explain warehouse as integrated source for mining
    • Discuss ETL, OLAP and knowledge discovery flow
    • Highlight decision support role in data science