Machine Learning Projects

A collection of machine learning algorithms implemented from scratch in Python, demonstrating fundamental ML concepts without relying on high-level libraries.

Projects Overview

1. Genetic Algorithm Solver

File: Genetic-Algorithm/GA.py

• Solves linear equation: a + 2b + 3c + 4d = 30
• Implements roulette wheel selection for parent selection
• Features crossover and mutation operators
• Interactive console for experimenting with genetic operations
• Tracks fitness scores and convergence across generations

2. Multiple Linear Regression

File: Multiple-Regression/Multiple-Regression.py

• Built using NumPy for matrix operations
• Predicts BMI categories from height and weight
• Manually computes regression coefficients (b₀, b₁, b₂)
• No sklearn or other ML libraries used
• Includes BMI dataset for training

3. Multi-Layer Perceptron (MLP)

File: MLP/Multi-Layer-Perceptron.py

• 3-5-1 architecture neural network
• Sigmoid activation function
• Manual backpropagation implementation
• Gradient calculation and weight updates from scratch
• Trains on BMI prediction task

4. Logic Gate Classification

Files:

• Logic-Gate-Classification/2input-AND-Function.py

• Logic-Gate-Classification/2input-OR-Function.py

• Logic-Gate-Classification/Sigmoid-Function.py

• Simple perceptron implementation

• Delta learning rule for weight updates

• Step and sigmoid activation functions

• Models AND/OR gates behavior

Key Features

• From Scratch Implementation: All algorithms coded without ML libraries
• Educational Focus: Clear code structure for learning purposes
• Real-World Applications: BMI prediction and classification tasks
• Interactive Elements: User input and experimentation options

Requirements

numpy
pandas

Usage

Each project can be run independently:

# Genetic Algorithm
python Genetic-Algorithm/GA.py

# Multiple Regression
python Multiple-Regression/Multiple-Regression.py

# MLP
python MLP/Multi-Layer-Perceptron.py

# Logic Gates
python Logic-Gate-Classification/2input-AND-Function.py
python Logic-Gate-Classification/2input-OR-Function.py

Learning Outcomes

• Understanding of evolutionary algorithms

• Linear regression mathematics

• Neural network fundamentals

• Gradient descent and backpropagation

• Activation functions and their applications