SPPU · Final Year Artificial Intelligence and Data Science · 2020 Course
Machine Learning
417521 · Semester VII · 3 credits · In-semester 30 + end-semester 70
Unit 1: Introduction to Machine Learning
6 hoursML versus programming, AI and data science; descriptive and predictive tasks; supervised, unsupervised, semi-supervised and reinforcement learning; geometric, probabilistic, logical, grouping and grading, parametric and non-parametric models; PCA and LDA.
Marks: 15 in the in-semester exam
- What machine learning is
- Types of learning
- Models of machine learning
- Principal component analysis
- Linear discriminant analysis
Unit 2: Regression
6 hoursRegression versus correlation; univariate, multivariate, linear and non-linear regression; bias–variance, over- and underfitting; polynomial, stepwise, tree, forest, support vector, ridge, lasso, elastic net and Bayesian regression; MSE, MAE, RMSE, R² and adjusted R².
Marks: 15 in the in-semester exam
- Simple and multiple linear regression
- Polynomial regression
- Tree-based and support vector regression
- Bias-variance trade-off
- Ridge and lasso regularisation
- Elastic net and Bayesian linear regression
- Regression metrics
Unit 3: Classification
6 hoursBinary and multiclass, balanced and imbalanced problems; linear classifiers; confusion matrix, accuracy, precision, recall and F-measures; one-vs-one and one-vs-all; per-class metrics; KNN; linear and soft-margin SVMs; RBF, polynomial and sigmoid kernels.
Marks: 17–18 in the end-semester exam
- Binary and multiclass classification
- Imbalanced classification
- Classification metrics
- Multiclass strategies and averaging
- K-nearest neighbours
- Support vector machines
- Kernel functions
Unit 4: Clustering
6 hoursAgglomerative and divisive hierarchical clustering; k-means, the elbow and silhouette methods, k-medoids and k-prototype; DBSCAN; Gaussian mixtures; applications in segmentation and anomaly detection.
Marks: 17–18 in the end-semester exam
- Hierarchical clustering
- K-means and choosing k
- K-medoids and K-prototype
- DBSCAN
- Gaussian mixture models
Unit 5: Ensemble Learning
6 hoursHomogeneous and heterogeneous ensembles; max, average and weighted voting; bagging and bootstrapping; AdaBoost, gradient boosting and XGBoost; stacking and blending; random forests.
Marks: 17–18 in the end-semester exam
- Voting ensembles
- Bagging and random forests
- Boosting and stacking
- XGBoost
Unit 6: Reinforcement Learning
6 hoursNeed for and elements of RL; Markov property, chains, reward processes and MDPs; return, policy and value functions; the Bellman equation; Q-tables, Q-functions and Q-learning.
Marks: 17–18 in the end-semester exam
- Elements of reinforcement learning
- Markov decision processes
- Bellman equation
- Q-learning
Unit contents are summarised from the official SPPU syllabus . AiCanCode.org is not affiliated with Savitribai Phule Pune University.