SUK · Third Year B.Tech. Computer Science and Engineering (Artificial Intelligence and Machine Learning) · NEP 2020
Machine Learning using Advanced Python
AIML3102 · Semester V · 2 credits · ISE 10 + mid-semester 30 + ESE 60
A 2-credit course with five units.
Unit 1: Introduction to Machine Learning
4 hoursDefinition, terminology, types and problem categories; the ML process and life cycle; generalisation, overfitting and underfitting; feature sets, extraction, subset selection, transformation and dimensionality reduction.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- What machine learning is
- Types of learning
- Bias-variance trade-off
- Feature selection: filter, wrapper and embedded methods
- Principal component analysis
Unit 2: Supervised Learning: Classification
6 hoursRegression versus classification; the confusion matrix, accuracy, precision, recall, ROC and F1; k-fold and stratified cross-validation; k-NN, naive Bayes and Bayesian networks; decision trees with ID3, Gini and CART; random forests; SVMs and the kernel trick.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Classification metrics
- Cross-validation and model evaluation
- K-nearest neighbours
- Bayesian learning and naive Bayes
- Decision tree learning and overfitting
- Bagging and random forests
- Support vector machines
Unit 3: Supervised Learning: Regression
6 hoursRegression models, distributions, hypothesis tests and confidence intervals; least squares, MSE, MAE and R²; gradient descent for linear regression; multivariate and polynomial regression; regularisation; maximum likelihood; logistic regression.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Hypothesis tests, significance and confidence intervals
- Simple and multiple linear regression
- Regression metrics
- Polynomial regression
- Ridge and lasso regularisation
- Logistic regression
Unit 4: Unsupervised Learning
4 hoursApplications and types of unsupervised learning; k-means and hierarchical clustering; the Apriori algorithm; the Weka and KEEL tools.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- K-means and choosing k
- Hierarchical clustering
- Association rules and Apriori
Unit 5: Neural Networks and ANN Algorithms
4 hoursThe neuron model and hypothesis; gradient descent for a neuron; neural networks and forward and backward propagation.
Examined in the 60-mark end-semester paper, and in the mid-semester exam if it is taught by then
- Perceptron and multilayer networks
- Backpropagation in neural networks
Unit contents are summarised from the official SUK syllabus . AiCanCode.org is not affiliated with Shivaji University, Kolhapur.