SGBAU · Third Year B.E. Electronics & Telecommunication Engineering · NEP 2020
Introduction to Machine Learning
6ET235PE (ii) · Elective III · Semester VI · 3 credits · IE 40 + ESE 60
One of the Elective III options.
Unit 1: Introduction to Machine Learning
6 hoursWhat machine learning is, why it is needed and where it is used; supervised, unsupervised and reinforcement learning; hypothesis space and inductive bias; training, testing, performance measures and cross-validation; over- and under-fitting and Occam's razor.
Examined in the 60-mark end-semester paper
- Introduction to machine learning
- Hypothesis space and inductive bias
- Cross-validation and micro and macro averaging
- Regression basics and the bias–variance trade-off
Unit 2: Linear Models and Regression
6 hoursLinear and multivariate regression, gradient descent, and the MSE, RMSE and R² metrics, with applications to sensors and signals.
Examined in the 60-mark end-semester paper
- Linear and logistic regression
- Gradient descent
- Regression evaluation metrics
Unit 3: Classification Techniques
6 hoursLogistic regression, k-NN, decision trees and random forests; accuracy, precision, recall and F1-score; applications in modulation recognition and fault detection.
Examined in the 60-mark end-semester paper
- K-nearest neighbours
- Decision trees and random forests
- Confusion matrix, precision, recall and F1
Unit 4: Unsupervised Learning
6 hoursK-means and hierarchical clustering, PCA and anomaly detection, with applications in spectrum sensing and network-traffic analysis.
Examined in the 60-mark end-semester paper
- K-means and K-medoids
- Clustering and hierarchical clustering
- PCA and LDA
- Outlier detection methods
Unit 5: Neural Networks and Deep Learning Basics
6 hoursThe perceptron learning algorithm, multilayer networks and backpropagation, activation functions and optimisation techniques; an introduction to deep neural networks and their applications.
Examined in the 60-mark end-semester paper
- Single-layer perceptron
- Multilayer perceptron
- Error backpropagation
- Activation functions
- Optimisers: AdaGrad, RMSProp and Adam
- Deep feedforward networks
Unit 6: Practical Considerations and Emerging Topics
6 hoursData pre-processing and feature engineering, over- and under-fitting, ML on embedded and IoT devices, reinforcement, edge and federated learning, and case studies.
Examined in the 60-mark end-semester paper
- Feature engineering
- Overfitting, dropout and regularisation
- Edge AI and TinyML
- Recent trends: self-supervised and federated learning, XAI and AutoML
Unit contents are summarised from the official SGBAU syllabus . AiCanCode.org is not affiliated with Sant Gadge Baba Amravati University (SGBAU).