Machine Learning

A collection of machine learning algorithms implemented in MATLAB/Octave and Python, covering supervised learning, unsupervised learning, and recommender systems.

Projects

#	Topic	Description	Language
01	Linear Regression	Univariate and multivariate regression with gradient descent and normal equations	MATLAB
02	Logistic Regression	Binary classification with regularization and polynomial decision boundaries	MATLAB
03	Multiclass Classification	One-vs-all logistic regression and pre-trained neural network for digit recognition	MATLAB
04	Neural Networks	Two-layer neural network trained from scratch with backpropagation	MATLAB
05	Bias-Variance Tradeoff	Polynomial regression with learning curves and regularization parameter selection	MATLAB
06	Support Vector Machines	Linear and RBF kernel SVMs, plus a spam email classifier	MATLAB
07	K-Means Clustering	Clustering algorithm applied to 2D data and image compression	MATLAB
08	Principal Component Analysis	Dimensionality reduction on 2D data, face images, and pixel data	MATLAB
09	Anomaly Detection	Gaussian-based anomaly detection for server monitoring	MATLAB
10	Recommender Systems	Collaborative filtering for movie recommendations (MovieLens dataset)	MATLAB
11	Normal Distribution	Gaussian PDF utilities, visualization, and maximum likelihood estimation	MATLAB
12	Linear Regression (Python)	Linear regression from scratch in Python with gradient descent	Python

Highlights

Supervised Learning

• Regression: Linear regression (univariate/multivariate), polynomial regression, regularization
• Classification: Logistic regression, one-vs-all, neural networks, SVMs, spam classification

Unsupervised Learning

• Clustering: K-Means with image compression application
• Dimensionality Reduction: PCA with eigenfaces and data visualization
• Anomaly Detection: Gaussian-based outlier identification

Recommender Systems

• Collaborative filtering for movie recommendations

Sample Visualizations

Linear Regression	Logistic Regression	K-Means Clustering

SVM Classification	PCA	Anomaly Detection

Credits

The MATLAB exercises are based on Andrew Ng's Machine Learning course on Coursera (Stanford University / DeepLearning.ai). All exercises were completed by Keivan Hassani Monfared.

License

This project is licensed under the GNU General Public License v3.0 - see the LICENSE file for details.