This repository houses the assignments and projects for the Machine Learning Course 2024 at K. N. Toosi University of Technology, under the supervision of Dr. Mahdi Aliyari, Associate Professor of Control and Mechatronics Engineering. It includes a variety of practical tasks and comprehensive projects designed to enhance your understanding of machine learning concepts. For detailed information on the topics covered in the course, please visit the Course webpage.
| Topic | Description | Access |
|---|---|---|
| McCulloch-Pitts Model | Implementing the McCulloch-Pitts neuron class with functionalities for step, sign, and ReLU activation functions. |  |
| Logistic Regression | Implementing Logistic Regression and SGD Classifier on synthetic and CWRU Bearing dataset. | |
| Linear Regression | Building regression models to predict specific weather-related outcomes on the Weather in Szeged 2006-2016 dataset. | |
| MLP Network | Utilize a multilayer perceptron network to classify faults in bearings. The model leverages features extracted from vibration data acquired from the CWRU Bearing dataset. | |
| Tree-Based Models | Apply Decision trees and Random Forest classifiers for a drug classification task with the help of pruning methods using the Drugs dataset. | |
| Bayes Model | Implementing a Naive Bayes classifier on the Heart Disease dataset to predict the presence of heart disease in patients. | |
| SVM Models | Using Support Vector Machine for classification task on Iris dataset with the polynomial kernel. | |
| Dimension Reduction | Apply linear and nonlinear dimension reduction techniques such as PCA, LDA, and t-SNE on the Iris dataset. | |
| Autoencoder Network | Training a Denoising Autoencoder (DAE) and neural network for denoising and classification tasks on Credit Card Fraud Detection dataset. | |
| Deep Q-Networks | Apply two popular Reinforcement Learning algorithms, DQN and DDQN, to the Lunar Lander environment. | |
Note
This table highlights the main topics such as implemented models and algorithms. However, each project employs various techniques for data preprocessing, visualization, and analysis. For more details, please refer to the directory of each mini-project.
For any questions or additional information regarding the Machine Learning Course 2024, please do not hesitate to contact me at mm.ghorbani@email.kntu.ac.ir.