This repository contains all computer assignments developed for the course
Machine Learning, offered by the
Electrical and Computer Engineering Department,
University of Tehran – Fall 1404,
under the supervision of Mohammadreza A. Dehaqani, Babak N. Arabi, and Mostafa Tavassolipour.
The projects in this course explore fundamental and advanced concepts of machine learning. They cover probabilistic classification, optimization techniques, supervised learning with SVMs and Decision Trees, unsupervised manifold learning, and deep neural architectures, culminating in a complete Spoken Language Identification system.
| CA | Project Title | Description | Link |
|---|---|---|---|
| CA1 | Machine Learning Basics | Foundational Bayesian theory, Maximum Likelihood Estimation (MLE), and Naive Bayes text/image classification. | View |
| CA2 | Optimization Techniques in Machine Learning | First and second-order optimization methods, SVM dual formulations, and Genetic Algorithm puzzle solvers. | View |
| CA3 | Support Vector Machines & Decision Trees | Implementation of ID3 decision trees from scratch and solving soft-margin SVMs via Quadratic Programming. | View |
| CA4 | Clustering, Dimensionality Reduction & Feature Selection | Unsupervised learning using K-Means, GMMs, and dimensionality reduction via PCA, Kernel PCA, and LDA. | View |
| CA5 | Deep Learning Foundations & Neural Architectures | Theoretical analysis of ReLU stability, Xavier initialization, and Transfer Learning with VGG-11 on CIFAR-10. | View |
| Final Project | Spoken Language Identification | An end-to-end pipeline for identifying spoken languages through acoustic feature engineering and classification. | View |
This repository is licensed under the MIT License.