UNIT-I: Introduction to Machine Learning
Introduction to Machine Learning – Types of learning: supervised, unsupervised and semi-supervised learning
– Hypothesis space and inductive bias – Training and test datasets – Cross validation – Overfitting and underfitting
– Bias and variance – Linear Regression: simple and multiple linear regression – Polynomial regression –
Evaluation metrics: accuracy, precision, recall, F1-score, MSE and RM – Feature scaling and normalization.
PART A(10x2=20)
1) Explain feature scaling, and why is normalization important before training a machine learning model?
2) Define Machine Learning.
3) Explain supervised learning? Give one example.
4) Define polynomial regression and also define best fit line.
5) Write in detail about cross-validation.
6) Differentiate between precision and recall.
7) Define bias in Machine Learning models.
8) Elaborate hypothesis space.
9) Define linear regression.
10) Illustrate underfitting.
11) List any two real-world applications of Machine Learning.
12) Explain the term inductive bias with an example.
13) Explain the term hypothesis in Machine Learning?
14) Elaborate unsupervised learning.
15) Define Mean Squared Error (MSE) and Root Mean Squared Error (RMSE).
16) State the bias–variance trade-off.
17) Give the correct explanation for simple linear regression?
18) Explain about multiple linear regression?
19) Illustrate Machine Learning.
20) Discuss about reinforcement learning.
21) Define test dataset.
22) Elaborate overfitting.
23) What is the F1-score, and why is it used instead of accuracy alone?
24) Discuss the training data with a neat example.
25) Define accuracy as an evaluation metric for a classification model.
PART-B ( 13 - Marks)
1) Analyze the bias–variance trade-off in Machine Learning.
2) Explain feature scaling and normalization. Describe common
techniques (Min-Max scaling, Standardization/Z-score) and discuss
why they are important for ML algorithms.
3) Describe different types of Machine Learning with suitable
examples.
4) Design a suitable evaluation strategy for a Machine Learning
model.
5) Explain simple linear regression in detail. Derive the normal
equations for estimating the regression coefficients using the least
squares method.
6) Analyze the role of inductive bias in improving model
generalization.
7) Discuss the need for model evaluation in Machine Learning.
8) Evaluate the performance of a regression model using suitable
metrics.
9) Explain training set, test set, and validation set with a neat diagram.
10) Compare supervised, unsupervised, and reinforcement learning.
11) Analyze a case where linear regression fails and explain why.
12) Explain multiple linear regression. Describe its mathematical
formulation, assumptions, and the method used to estimate the
regression coefficients.
13) Develop a Machine Learning workflow from data collection to
evaluation.
14) Critically analyze the limitations of linear regression models.
15) Illustrate a real-world application of Machine Learning using
regression.
16) Explain the concept of hypothesis space and inductive bias in detail.
17) Analyze the impact of model complexity on regression
performance.
18) Explain cross-validation techniques in detail.
19) Propose a Machine Learning approach to solve a prediction
problem.
20) Elaborate Machine Learning and its importance in modern
computing.
21) Analyze how cross-validation helps prevent overfitting.
22) Explain the evaluation metrics accuracy, precision, recall, and F1-
score for classification problems with appropriate formulas and a
suitable example using a confusion matrix.
23) Compare overfitting and underfitting in Machine Learning models.
24) Design a regression model for a real-world problem and justify the
choice.
PART-C (15 - Marks)
1 Describe the complete workflow of a machine learning system. Include
data splitting, model training, evaluation, and validation
2 Explain polynomial regression and its significance.
3 Discuss methods used to evaluate regression models.
4 Explain Mean Squared Error (MSE) and R² score.
5 Explain multiple linear regression with an example.
6 Discuss the evaluation metrics MSE and RMSE for regression problems
and accuracy, precision, recall, and F1-score for classification problems.
Explain when each metric is appropriate, with examples.
7 Explain simple and multiple linear regression in detail. Derive the normal
equations for both cases and compare their assumptions, advantages, and
limitation