Chapter 3: Basics of Machine Learning - 20 Questions
Q: Define Machine Learning and explain its significance in AI.
A: Machine Learning (ML) is a subset of AI that enables systems to learn from data without explicit
programming. Significance includes automation of decision-making, pattern recognition, and AI applications
like speech recognition.
Q: Differentiate between Supervised, Unsupervised, and Reinforcement Learning with
examples.
A: Supervised Learning: Trained on labeled data (e.g., Email spam detection). Unsupervised Learning:
Identifies patterns in unlabeled data (e.g., Customer segmentation). Reinforcement Learning: Learns via
rewards and penalties (e.g., Chess-playing AI).
Q: What are the key components of a Machine Learning model?
A: Dataset, Features, Model, Loss Function, Optimizer, and Evaluation Metrics.
Q: Explain the concepts of training, validation, and testing datasets in Machine Learning.
A: Training: Used to train the model. Validation: Tunes hyperparameters. Testing: Evaluates final model
performance.
Q: What is overfitting and underfitting? How can they be prevented?
A: Overfitting: Model learns noise instead of pattern (High training accuracy, low test accuracy). Underfitting:
Model is too simple (Poor accuracy). Prevention: More data, regularization, cross-validation, simpler model.
Q: Differentiate between classification and regression with suitable examples.
A: Classification: Predicts categorical output (e.g., Spam detection). Regression: Predicts continuous output
(e.g., House price prediction).
Q: Apply Linear Regression to predict the test score for a student who studied for 5 hours.
A: Using Linear Regression on the dataset, the predicted score for 5 hours is 65.
Q: Implement K-Means Clustering to group customers based on spending habits.
A: Applying K-Means clustering with k=2, customers are grouped based on income and spending score.
Q: Compute accuracy, precision, recall, and F1-score for the given confusion matrix.
A: Accuracy = 0.85, Precision = 0.80, Recall = 0.89, F1-score = 0.84.
Q: Implement a Decision Tree classifier to predict whether a customer will buy a product.
A: Using a Decision Tree on the dataset, a customer with Age=40, Salary=60000 is predicted to buy the
product.
Q: What are feature scaling techniques, and why are they important in Machine Learning?
A: Feature scaling techniques include Normalization and Standardization. They ensure that features
contribute equally to the model and improve convergence in optimization algorithms like Gradient Descent.
Q: Explain the concept of bias-variance tradeoff in Machine Learning.
A: High bias leads to underfitting (too simple model), while high variance leads to overfitting (too complex
model). The goal is to find a balance between the two for optimal performance.
Q: What are hyperparameters in Machine Learning? Give examples.
A: Hyperparameters are configuration settings that are set before training the model, such as learning rate,
number of hidden layers in a neural network, and the number of clusters in K-Means.
Q: What is cross-validation, and why is it used?
A: Cross-validation is a technique to evaluate a model's performance by dividing the dataset into training and
testing subsets multiple times. It helps in reducing overfitting and ensures better generalization.
Q: Explain the working of the Random Forest algorithm.
A: Random Forest is an ensemble learning method that creates multiple decision trees and averages their
predictions to improve accuracy and reduce overfitting.
Q: What is Principal Component Analysis (PCA) in Machine Learning?
A: PCA is a dimensionality reduction technique that transforms data into a set of orthogonal components to
reduce redundancy while retaining important information.
Q: Differentiate between Bagging and Boosting in ensemble learning.
A: Bagging (e.g., Random Forest) reduces variance by training multiple models in parallel. Boosting (e.g.,
AdaBoost, XGBoost) reduces bias by training models sequentially, giving more weight to misclassified
samples.
Q: Explain the role of activation functions in Neural Networks.
A: Activation functions introduce non-linearity into neural networks, enabling them to learn complex patterns.
Common functions include ReLU, Sigmoid, and Tanh.
Q: What is Gradient Descent, and how does it work?
A: Gradient Descent is an optimization algorithm that minimizes the loss function by iteratively updating
model parameters in the direction of the negative gradient.
Q: Explain the difference between Content-Based and Collaborative Filtering in
recommendation systems.
A: Content-Based Filtering recommends items based on item features and user preferences, while
Collaborative Filtering relies on user-item interactions and similarities between users or items.
