AD4501 - Machine Learning Department of ADS 2025-2026

AD4501: MACHINE LEARNING

Corresponding Lab, with code (If any): - AD4509 - MACHINE LEARNING LABORATORY
Course Prerequisites: -
Course Outcomes
On successful completion of this course, the student will be able to
C301.1 Gain knowledge about basic concepts of machine learning techniques
C301.2 Develop predictive model based on both input and output data
C301.3 Construct supervised learning models.
C301.4 Construct unsupervised learning algorithms
C301.5 Evaluate and compare different model.

MAPPING BETWEEN CO AND PO, PSO WITH CORRELATION LEVEL 1/2/3

Pos PSOs
AD4501
PO1 PO2 PO3 PO4 PO5 PO6 PO7 PO8 PO9 PO10 PO11 PO12 PSO1 PSO2 PSO3 PSO4
C301.1 3 3 3 3 2 2 2 2 2 2 2 2 3 3 2 2
C301.2 3 3 3 3 2 2 2 2 2 2 2 2 3 3 2 2
C301.3 3 3 3 3 2 2 2 2 2 2 2 2 3 3 2 2
C301.4 3 3 3 3 2 2 2 2 2 2 2 2 3 3 2 2
C301.5 3 3 3 3 2 2 2 2 2 2 2 2 3 3 2 2
RELATION BETWEEN COURSE CONTENT WITH COs
UNIT I- INTRODUCTION
Knowledge No. of Hrs to Books
S. No Topic
level be handled Referred
1. Machine Learning L2 1 T1
2. Types of Machine Learning L2 2 T1
3. Supervised Learning L2 1 T1
4. Unsupervised Learning L2 1 T1
5. Basic Concepts in Machine Learning L2 2 T1
6. Machine Learning Process L1,L2 1 T1
7. Weight Space L2 1 T1
8. Testing Machine Learning Algorithms L2 1 T1
9. A Brief Review of Probability Theory L3 1 T1
10. Turning Data into Probabilities L2 1 T1
11. The Bias-Variance Trade-off L3 2 T1
12. FIND–S Algorithm L2 1 T1
13. Candidate Elimination Algorithm L3 1 T1

UNIT II- SUPERVISED LEARNING

Knowledge No. of Hrs to Books
S. No Topic
level be handled Referred
1. Linear Models for Regression L1,L2 1 T1
2. Linear Basis Function Models L2 1 T1
3. The Bias-Variance Decomposition L2 2 T1
4. Bayesian Linear Regression L2 2 T1
5. Common Regression Algorithms L2 1 T1
6. Simple Linear Regression L2 1 T1
7. Multiple Linear Regression L2 1 T1
8. Linear Models for Classification L2 1 T1
9. Discriminant Functions L2 1 T1
10. Probabilistic Generative Models L3 1 T1
11. Probabilistic Discriminative Models L3 2 T1
12. Laplace Approximation L3 1 T1

St. Joseph’s Institute of Technology

Page 1 of 10
 AD4501 - Machine Learning Department of ADS 2025-2026
13. Bayesian Logistic Regression L2 1 T1
14. Common Classification Algorithms L3 1 T1
15. k-Nearest Neighbors L3 2 T1
16. Random Forest model L3 2 T1
17. Support Vector Machines L3 1 T1

UNIT III- UNSUPERVISED LEARNING

Knowledge No. of Hrs to Books
S. No Topic
level be handled Referred
1. Mixture Models and EM L2 1 T1
2. K-Means Clustering L1,L2 1 T1
3. Dirichlet Process Mixture Models L2 1 T1
4. Spectral Clustering L1,L2,L3 1 T1
5. Hierarchical Clustering L2 2 T1
6. The Curse of Dimensionality L2 1 T1
7. Dimensionality Reduction L2 1 T1
8. Principal Component Analysis L2 1 T1
9. Latent Variable Models (LVM) L1,L2 1 T1
10. Latent Dirichlet Allocation (LDA) L3 1 T1

UNIT IV- NEURAL NETWORKS

Knowledge No. of Hrs to Books
S. No Topic
level be handled Referred
1. Multilayer perceptron L1,L2 2 T1
2. Activation functions L2 1 T1
3. Network training L2 1 T1
4. Gradient descent optimization L2 1 T1
5. Stochastic gradient descent L2 1 T1
6. Error backpropagation L1,L2 1 T1
7. From shallow networks to deep networks L2 1 T1
8. Unit saturation (aka the vanishing gradient L2 T1
1
problem)
9. ReLU, hyper parameter tuning L1 1 T1
10. Batch normalization L2 1 T1
11. Regularization L1 2 T1
12. Dropout L3 1 T1

UNIT V- DESIGN AND ANALYSIS OF MACHINE LEARNING EXPERIMENTS

Knowledge No. of Hrs to Books
S. No Topic
level be handled Referred
1. Guidelines for machine learning experiments L2 1 T1
2. Cross Validation (CV) and resampling L2 2 T1
3. K-fold CV L2 1 T1
4. bootstrapping L2 1 T1
5. measuring classifier performance L2 1 T1
6. assessing a single classification algorithm L2 1 T1
7. T test L2 2 T1
8. McNemar’s test L2 1 T1
9. K-fold CV paired t test L2 1 T1

L1- Remember; L2- Understand; L3- Apply; L4- Analyze; L5- Evaluate; L6- Create

St. Joseph’s Institute of Technology

Page 2 of 10
 AD4501 - Machine Learning Department of ADS 2025-2026

TEXT BOOK:
1. Ethem Alpaydin, “Introduction to Machine Learning”, Fourth Edition, The MIT Press, 2020.
2. Stephen Marsland, “Machine Learning: An Algorithmic Perspective, “Second Edition”, CRC Press, 2018.

REFERENCES:
1. Christopher Bishop, “Pattern Recognition and Machine Learning”, Springer, 2006.
2. Tom Mitchell, “Machine Learning”, McGraw-Hill, 2017.
3. Mehryar Mohri, Afshin Rostamizadeh, Ameet Talwalkar, “Foundations of Machine Learning”, 2nd Edition,MIT
Press, 2012,
4. Ian Goodfellow, Yoshua Bengio, Aaron Courville, “Deep Learning”, MIT Press, 2016
5. Sebastain Raschka, Vahid Mirjalili , “Python Machine Learning”, Packt publishing, 3rd Edition,2019.

UNIT I INTRODUCTION
Machine Learning – Types of Machine Learning – Supervised Learning – Unsupervised Learning – Basic Concepts in
Machine Learning – Machine Learning Process – Weight Space – Testing Machine Learning Algorithms – A Brief Review
of Probability Theory –Turning Data into Probabilities – The Bias-Variance Trade-off, FIND–S Algorithm, Candidate
Elimination Algorithm

PART-A CO Knowledge
Statement Level((BL1/BL2/BL
3/ BL4/ BL5/BL6)
1 Define machine learning. C301.1 BL2
2 Why Is Machine Learning Important? C301.1 BL1
3 List some disciplines and examples of their influence on machine learning C301.1 BL2
4 What is the difference between artificial intelligence and machine learning C301.1 BL2
methods?
5 List the types of machine learning. C301.1 BL1
6 What is supervised learning? C301.1 BL2
7 What is unsupervised learning? C301.1 BL1
8 What is reinforcement learning? C301.1 BL1
9 What is evolutionary learning? C301.1 BL1
10 Differentiate Parametric model and Non-parametric model. C301.1 BL2
11 Define the concept of No free lunch theorem. C301.1 BL2
12 List the steps involved in machine learning process. C301.1 BL2
13 What is activation function? C301.1 BL2
14 Why can’t we do it without activating the input signal? C301.1 BL1
15 List the various activation function. C301.1 BL1
16 What is meant by Overfitting C301.1 BL2
17 What is meant by Underfitting C301.1 BL2

18 Write about Confusion Matrix in Machine Learning C301.1 BL2

19 What is Bias-Variance Tradeoff in machine learning? C301.1 BL1,BL2
20 Define Version space. C301.1 BL2
21 List the advantages and disadvantages in Version space. C301.1 BL2
22 What is precision in machine learning?(Nov/Dec 2024) C301.1 BL1
23 What is recall in machine learning? C301.1 BL2
24 List out the classification of Machine Learning (April/May 2025) C301.1 BL2
25 Write the main objective of the FIND-S algorithm in machine C301.1 BL2
Learing.(April/May 2025)

St. Joseph’s Institute of Technology

Page 3 of 10
 AD4501 - Machine Learning Department of ADS 2025-2026
PART-B
1 a) List and explain some successful applications of machine C301.1 BL1, BL2
learning.
b) List some disciplines and examples of their influence on machine
learning.
2 Elaborate the Perspectives and issues in machine learning C301.1 BL1, BL2
3 Explain in detail about Basic Concepts in Machine Learning. C301.1 BL2,BL3
4 Explain Machine Learning Process in detail with an example application. C301.1 BL4
5 Explain in detail about Testing Machine Learning Algorithms. C301.1 BL2,BL3
6 Explain supervised learning with the concept of regression and C301.1 BL1,BL3
classification.
7 Explain the important statistical concepts in detail. C301.1 BL2,BL3
8 Explain the following C301.1 BL1,BL4
a) Turning data into probabilities b) BiasVarianceTradeoff
9 Write FIND-S algorithm. Trace the algorithm with example. C301.1 BL2,BL4
10 Explain version space and List-Then-Eliminate algorithm with an C301.1 BL2,BL4
example.
11 Explain CANDIDATE ELIMINATION learning algorithm with an C301.1 BL2,BL3
example.
12 List and explain the various remarks on version spaces and candidate C301.1 BL2,BL4
elimination.
13 Distinguish between supervised and unsupervised learning with suitable C301.1 BL4
examples.(April/May 2025)
14 Examine the candidate Elimination Algorithm with a worked-out C301.1 BL4
example.(April/May 2025)
UNIT II SUPERVISED LEARNING
Linear Models for Regression – Linear Basis Function Models – The Bias-Variance Decomposition – Bayesian Linear
Regression – Common Regression Algorithms – Simple Linear Regression – Multiple Linear Regression – Linear Models
for Classification – Discriminant Functions – Probabilistic Generative Models – Probabilistic Discriminative Models –
Laplace Approximation – Bayesian Logistic Regression – Common Classification Algorithms – k-Nearest Neighbors –
Decision Trees – Random Forest model – Support Vector Machines
PART-A CO Knowledge
Statement Level((BL1/BL2/BL
3/ BL4/ BL5/BL6)
1 What is linear regression? C301.2 BL1
2 State the assumptions in a linear regression model. C301.2 BL2
3 What is linear separability? C301.2 BL3
4 Express Linear Basis Function Models . C301.2 BL2
5 What is Gaussian basis function? C301.2 BL2
6 Why is Gaussian basis function model useful? C301.2 BL2
7 What is Bias-Variance Decomposition? C301.2 BL2
8 When to use bias-variance decomposition? C301.2 BL2
9 Define simple linear regression. C301.2 BL2
10 List the advantages and disadvantages of Bayesian Regression. C301.2 BL1,BL2
11 List the common regression algorithm in machine learning. C301.2 BL2
12 What is Bayesian Regression? C301.2 BL2
13 Define Multiple Linear Regression. C301.2 BL1
14 What is the idea of decision tree? C301.2 BL2
15 Differentiate Discriminative and Generative models. C301.2 BL1
16 What is Laplace approximation used for? C301.2 BL2
17 What is Bayesian logistic regression model? C301.2 BL2,BL3
18 What is k-NN algorithm? C301.2 BL2,BL4

St. Joseph’s Institute of Technology

Page 4 of 10
 AD4501 - Machine Learning Department of ADS 2025-2026
19 Define classification models and various classification models in machine C301.2 BL1,BL2
learning.
20 Define (information) Entropy C301.2 BL2
21 List the advantages and disadvantages of decision tree. C301.2 BL1
22 Write function for computing the entropy. C301.2 BL5
23 Define information gain. C301.2
24 List some decision tree algorithms. C301.2 BL2,BL3
25 Write note on ID3 algorithm. C301.2 BL1
26 What is Gini impurity? C301.2 BL1
27 What is CART? C301.2 BL1
28 What is a random forest in machine learning? C301.2 BL1
29 What is SVM? C301.2 BL1
30 List the various choices for basis function. C301.2 BL2
31 How deoes increasing model complexity affect bias and C301.2 BL2
variance?(April/May 2025)
32 Why are Random forests more robust than a single Decision C301.2 BL2
Tree?(April/May 2025)
UNIT-II / PART-B
1 Explain in detail about Linear Basis Function Models. C301.2 BL2,BL3
2 Discuss in detail about Bias-Variance Decomposition. C301.2 BL2,BL3
3 Write a brief note on Parameter distribution, Predictive distribution and C301.2 BL1,BL4
Equivalent kernel.
4 Explain in detail about Fisher’s linear discriminant and Fisher’s C301.2 BL2,BL3
discriminant for multiple classes.
5 Explain Perceptron learning algorithm with an example. C301.2 BL1,BL3
6 Explain Probabilistic Generative Models in detail. C301.2 BL1,BL3
7 Explain Probabilistic Discriminative Models in detail. C301.2 BL1,BL2
8 Write a brief note on the Laplace Approximation. C301.2 BL2,BL3
9 Explain Bayesian Logistic Regression in detail. C301.2 BL1,BL3
10 Explain K-nearest neighbor methods in detail. C301.2 BL2,BL4
11 Explain ID3 algorithm with example. C301.2 BL1,BL2
12 Explain classification and regression trees with an example. C301.2 BL1,BL4
13 Explain Random Forest algorithm with an example. C301.2 BL1,BL2
14 Explain support vector machine in detail. C301.2 BL2,BL4
15 Compare and Contrast Bayesian Linear Regression with ordinary Least C301.2 BL4
Squares (OLS) Regression. Provide an example where it can be effectively
applied in the real world.(April/May 2025)
16 Interpret the k-Nearest Neighbors with suitable example (April/May 2025) C301.2 BL4
UNIT III UNSUPERVISED LEARNING
Mixture Models and EM – K-Means Clustering – Dirichlet Process Mixture Models – Spectral Clustering – Hierarchical
Clustering – The Curse of Dimensionality – Dimensionality Reduction – Principal Component Analysis – Latent Variable
Models (LVM) – Latent Dirichlet Allocation (LDA)
UNIT-III / PART-A CO Knowledge Level
Statement Knowledge
Level((BL1/BL2/BL
3/ BL4/ BL5/BL6)
1 What is mixture models in machine learning? C301.3 BL1
2 What is a Gaussian mixture model used for? C301.3 BL2

3 List the two main applications of mixture models. C301.3 BL2

St. Joseph’s Institute of Technology

Page 5 of 10
 AD4501 - Machine Learning Department of ADS 2025-2026
4 What is expectation–maximization (EM) algorithm? C301.3 BL2
5 What is Convergence in the EM algorithm? C301.3 BL1
6 Differences between K-means and GMM Gaussian mixture model? C301.3 BL2
7 Define Dirichlet process. C301.3 BL2
8 Why do we use Dirichlet distribution? C301.3 BL2
9 Draw a Graphical Model for Finite Mixture Model with Dirichlet C301.3 BL2
Distribution
10 Why is it called spectral clustering? C301.3 BL2
11 What does the graph Laplacian do? C301.3 BL4
12 What is the curse of dimensionality reduction in machine learning? C301.3 BL1
13 Difference between k-means Clustering and Hierarchical Clustering. C301.3 BL2
14 List and explain two main approaches of hierarchical clustering. C301.3 BL2
15 List the two main algorithms in Dimensionality Reduction C301.3 BL2
16 When should I use PCA? C301.3 BL1
17 How the Linear Discriminant Analysis (LDA) work? C301.3 BL1,BL2
18 Why are latent variables introduced in the EM algorithm? C301.3 BL2,BL4
19 What does the E step do in the EM algorithm for Gaussian mixture models? C301.3 BL2
20 What is the clustering complexity of agglomerative and divisive clustering C301.3 BL2
during chaining phenomenon?
21 What is Gaussian mixture model? C301.3 BL2
22 List the different ways to do dimensionality reduction C301.3 BL2
23 Define feature selection. C301.3 BL1
24 Define feature derivation C301.3 BL2
25 What is Linear Discriminant Analysis (LDA)? C301.3 BL1
26 What is the idea of principal component analysis? C301.3 BL2
27 Define spectral decomposition. C301.3 BL2
28 Find the spectral decomposition for A = (3 2 2 3) C301.3 BL3
29 Differentiate Agglomerative and Divisive Hierarchical C301.3 BL2
Clustering?(April/May 2025)
30 What does PCA aim to achieve in dimensionality reduction? (April/May C301.3 BL2
2025)
UNIT-III / PART-B
1 Explain the k-Means Algorithm with an example. C301.3 BL1,BL2
2 Explain the concept of EM Algorithm. C301.3 BL2,BL4
3 Define clustering. What are the different types of clustering explain in C301.3 BL2,BL4
detail?
4 Give a detail note on Mixture models in machine Learning. C301.3 BL2,BL4
5 Explain the three different ways to do dimensionality reduction in detail. C301.3 BL1,BL3
6 Explain Linear Discriminant analysis in detail. C301.3 BL1,BL3
7 Explain principal component analysis algorithm in detail. C301.3 BL2,BL4
8 Use the LDA on the Iris dataset. Compare the results with using PCA, C301.3 BL1,BL3
which is not supervised and will not therefore be able to find the same
space.
9 Explain Latent Variable Models (LVM) in details C301.3 BL2
10 Explain in detail about Latent Dirichlet Allocation (LDA) C301.3 BL2
11 Describe the concept of Mixture Models and the role of the Expectation- C301.3 BL4
Maximization (EM) algorithm in estimating the parameters of a Gaussian
Mixture Model (GMM).(April/May 2025)
12 Discuss in detail about Principal Component Analysis (PCA) and how it C301.3 BL4
helps mitigate its effects.(April/May 2025)

St. Joseph’s Institute of Technology

Page 6 of 10
 AD4501 - Machine Learning Department of ADS 2025-2026
UNIT IV NEURAL NETWORKS
Multilayer perceptron, activation functions, network training – gradient descent optimization – stochastic gradient descent,
error backpropagation, from shallow networks to deep networks –Unit saturation (aka the vanishing gradient problem) –
ReLU, hyper parameter tuning, batch normalization, regularization, dropout.
UNIT IV - PART A CO Knowledge Level
Statement Knowledge
Level((BL1/BL2/BL
3/ BL4/ BL5/BL6)
1 What is a neural network? C301.4 BL1
2 What are the components of neural network? C301.4 BL2
3 Define multilayer perceptron (MLP) C301.4 BL2
4 What are the layers in multilayer perceptron (MLP) C301.4 BL2
5 What is the function of Input layer? C301.4 BL2
6 Define the functions of Hidden layers? C301.4 BL2
7 What is the function of output layer? C301.4 BL2
8 What is Activation functions? C301.4 BL2
9 List some of the Activation functions C301.4 BL2
10 Define network training: C301.4 BL1
11 What is Gradient descent optimization? C301.4 BL1
13 What is Loss function? C301.4 BL1
14 Define stochastic gradient descent: C301.4 BL2
15 What is Error backpropagation C301.4 BL2
16 What is the use of Back propagation? C301.4 BL3
17 Define Shallow Networks C301.4 BL2
18 What is the limitations of Shallow Networks C301.4 BL3
19 What is Deep Networks C301.4 BL1
20 What are the advantages of Deep Networks C301.4 BL1
21 Define Unit saturation: C301.4 BL2
22 Define Output Saturation: C301.4 BL2
23 Define ReLU C301.4 BL1
24 What is Hyper parameter tuning ? C301.4 BL1
25 Define Batch normalization: C301.4 BL1
26 What is Regularization: C301.4 BL1
27 Outline the role of hidden layers in a multilayer C301.3 BL2
perceptron(MLP)(April/May 2025)
28 How does the learning rate affect gradient descent optimization? C301.3 BL2
(April/May 2025)
UNIT IV - PART B
1 Explain in detail about Multilayer perceptron C301.4 BL1,BL2
2 Illustrate in detail about activation functions C301.4 BL2,BL3
3 Explain in detail about network training and what are its advantageous C301.4 BL2,BL3
4 Explain in detail about gradient descent optimization algorithm C301.4 BL1,BL2
5 Illustrate in details about stochastic gradient descent algorithms C301.4 BL2,BL3
6 Explain in detail about error back propagation methods C301.4 BL2,BL3
8 Explain in detail about from shallow networks to deep networks C301.4 BL2,BL3
7 Illustrate in detail about Unit saturation methods C301.4 BL2,BL3
8 Write short notes on ReLU and hyper parameter tuning methods C301.4 BL1,BL2
9 Write short notes on batch normalization, regularization and dropout C301.4 BL1,BL3

St. Joseph’s Institute of Technology

Page 7 of 10
 AD4501 - Machine Learning Department of ADS 2025-2026
10 Using the X-ray images create a neural network model for lung Cancer C301.4 BL6
detection
11 Explain the architecture of a Multilayer Perceptron (MLP) and the role of C301.3 BL4
activation functions in neural networks.(April/May 2025)
12 Relate how the dropout technique works and analyze its effect on C301.3 BL4
network training.(April/May 2025)
13 Design a systematic approach to tuning hyper parameters and evaluate C301.3 BL5
different tuning methods.(April/May 2025)

UNIT – V DESIGN AND ANALYSIS OF MACHINE LEARNING EXPERIMENTS

Guidelines for machine learning experiments, Cross Validation (CV) and resampling – K-fold CV, bootstrapping,
measuring classifier performance, assessing a single classification algorithm and comparing two classification algorithms
– t test, McNemar’s test, K-fold CV paired t test
UNIT V - PART A CO Knowledge Level
Statement Knowledge
Level((BL1/BL2/BL
3/ BL4/ BL5/BL6)
1 What is the importance of experimental design in machine learning C301.5 BL2
research?
2 Explain the components of a typical machine learning experiment design. C301.5 BL2
3 How does the choice of evaluation metrics impact the design of machine C301.5 BL1
learning experiments?
4 Discuss the significance of cross-validation in the design of machine C301.5 BL2
learning experiments.
5 What are the common strategies for dataset splitting in machine learning C301.5 BL2
experiments, and when are they used?
6 Explain the concept of hyperparameter tuning and its role in optimizing C301.5 BL1
machine learning models.
7 How do experimental design considerations differ for supervised, C301.5 BL1
unsupervised, and reinforcement learning tasks?
8 Discuss the importance of reproducibility and replicability in machine C301.5 BL1
learning experiments.
10 What are some common pitfalls to avoid when designing and conducting C301.5 BL2
machine learning experiments?
11 Explain the concept of experimental bias and its impact on the validity of C301.5 BL1
machine learning results.
12 How can experimental design techniques such as factorial design and C301.5 BL2
randomized control trials be applied to machine learning research?
13 Discuss the trade-offs between computational resources and experimental C301.5 BL2
complexity in machine learning research.
14 What role does experimental documentation and reporting play in ensuring C301.5 BL1
the transparency and reproducibility of machine learning research?
15 How can experimental design principles be adapted to address challenges C301.5 BL2
specific to real-world applications of machine learning, such as domain
shift and concept drift?
16 What are the ethical considerations involved in the design and conduct of C301.5 BL2
machine learning experiments, and how can they be addressed?
17 Explain the concept of experimental control and its role in establishing C301.5 BL1
causal relationships in machine learning research.
18 How does the choice of model evaluation strategy, such as holdout C301.5 BL1
validation, k-fold cross-validation, or bootstrapping, affect the reliability
of experimental results?

St. Joseph’s Institute of Technology

Page 8 of 10
 AD4501 - Machine Learning Department of ADS 2025-2026
19 Discuss the use of statistical hypothesis testing in the analysis of machine C301.5 BL2
learning experiments and its limitations.
20 How does cross-validation help in mitigating the risk of overfitting in C301.5 BL1
machine learning experiments?
21 What are the key considerations when selecting evaluation metrics for C301.5 BL1
machine learning experiments?
22 Discuss the trade-offs involved in selecting between holdout validation and C301.5 BL2
k-fold cross-validation for evaluating machine learning models.
23 What are some emerging trends and advancements in experimental design C301.5 BL2
and analysis techniques in machine learning research?
24 Discuss the role of experimental control in establishing causal C301.5 BL1
relationships in machine learning research.
25 Demonstrate Bootstrapping.(April/May 2025) C301.3 BL2
26 Why should we use multiple evaluation metrics instead of relying only on C301.3 BL2
accuracy?(April/May 2025)
UNIT-V / PART-B
1 Explain in detail the concept of experimental control in machine learning C301.5 BL2,BL3
research and its significance in establishing causal relationships between
variables. Provide examples to illustrate the importance of experimental
control in ensuring the internal validity of experiments.
2 Discuss the challenges and considerations involved in designing machine C301.5 BL2,BL4
learning experiments with proper experimental control. How can
researchers identify and mitigate potential sources of bias and
confounding variables to ensure the reliability of causal inference?
3 Critically evaluate the role of randomized controlled trials (RCTs) in C301.5 BL2,BL4
establishing causal relationships in machine learning research. Compare
and contrast the strengths and limitations of RCTs with other
experimental designs, such as observational studies and quasi-
experimental designs, in the context of causal inference.
4 Analyze the impact of experimental design choices, such as sample size C301.5 BL2,BL3
determination, randomization procedures, and control group selection, on
the validity and generalizability of causal conclusions drawn from
machine learning experiments. How do these design choices affect the
robustness and reproducibility of experimental findings?
5 Explore the ethical considerations associated with experimental control in C301.5 BL2,BL3
machine learning research, particularly regarding the manipulation of
experimental conditions and the treatment of human subjects or sensitive
data. How can researchers ensure the ethical conduct of experiments
while maintaining rigorous experimental control?
6 Investigate the role of feature selection and variable manipulation C301.5 BL2,BL3
techniques in experimental control for machine learning experiments.
How do researchers determine which variables to include as independent
variables and how to manipulate them to establish causal relationships
effectively?
7 Examine the relationship between experimental control and external C301.5 BL2,BL3
validity in machine learning research. How do researchers balance the
need for internal validity through strict experimental control with the
desire to generalize findings to real-world contexts?
8 Evaluate the impact of contextual factors, such as dataset characteristics, C301.5 BL2,BL4
algorithmic complexity, and experimental context, on the effectiveness of
experimental control in machine learning research. How do researchers
account for these factors when designing and interpreting experiments?
9 Discuss the role of pre-processing and data cleaning techniques in C301.5 BL2,BL4
ensuring the validity of experimental control in machine learning
research. How do researchers address issues such as data contamination,

St. Joseph’s Institute of Technology

Page 9 of 10
 AD4501 - Machine Learning Department of ADS 2025-2026
missing values, and data imbalance to minimize the risk of bias and
confounding effects?
10 Reflect on the future directions and challenges in experimental control for C301.5 BL2,BL4
machine learning research, particularly in emerging areas such as deep
learning, reinforcement learning, and interpretability. How can
researchers innovate experimental designs and methodologies to address
complex causal relationships in increasingly sophisticated machine
learning models?
11 Evaluate K-fold Cross-Validation (K-fold CV) and bootstrapping. C301.3 BL4
Explain how these techniques work and compare their advantages and
limitations.(April/May 2025)
12 Compare and contrast the single classification algorithm and two C301.3 BL4
classification algorithm.(April/May 2025)
13 Categorize the Decision tree and Random forest tree. Give a suitable real C301.3 BL5
time example for each.(April/May 2025)

St. Joseph’s Institute of Technology

Page 10 of 10