Machine-Learning-Projects-MATLAB

Descrption

These projects are part of class assignments that were aimed to inculcate a deeper understanding of machine learning algorithms such as Classification with Perceptron, Support Vector Machine and Principal Component Analysis (PCA). These were implemented in MATLAB without using any libraries. Implementing the algorithms such as SVM, Stochastic Gradient Descent (SGD), and PCA from scratch enhanced the understanding of the core ML concepts like classiication, optimization and dimensionality reduction

Classification Models

1. Perceptron Learning

This project is an implementation of the linear perceptron using stochastic gradient descent (SGD) in MATLAB. The dataset is a set of points in 2D space which are linearly separable.

2. Support Vector Machine

This project implements Support Vector Machine algorithm to classify data using different kernels(linear, polynomia and RBF).

Error corresponding to different polynomial degrees

Principal Component Analysis

This project is an implementation of the PCA algorithm in machine learning using MATLAB. A dataset of teapot images is used and the images are re-created by using only three principal features of the images.

Before Applying PCA	After Applying PCA