Machine Learning Algorithms

This repository contains Jupyter Notebooks implementing various foundational machine learning algorithms. Each notebook is designed to be beginner-friendly, with detailed explanations and examples, making it suitable for both learning and experimentation.

Included Algorithms

1. K-Nearest Neighbors (KNN)

• Description: Implements the KNN algorithm, a non-parametric method used for classification and regression.
• Features:
  • Custom implementation using numpy.
  • Data visualization with matplotlib.
  • Designed to work with CSV datasets.
• Notebook: KNN Algorithm.ipynb
• More Info: KNN README

---

2. Linear Regression

• Description: Implements Linear Regression for predicting a target variable based on input features.
• Features:
  • Simple and customizable framework.
  • Data visualization with matplotlib.
  • Works with CSV datasets for easy reproducibility.
• Notebook: Linear Regression Algorithm.ipynb
• More Info: Linear Regression README

---

3. Logistic Regression

• Description: Implements Logistic Regression for binary classification tasks.
• Features:
  • Progress tracking using tqdm.
  • Visualization of results with matplotlib.
  • Utilizes numpy for computations.
• Notebook: Logistic Regression Algorithm.ipynb
• More Info: Logistic Regression README

---

4. Convolutional Neural Network (CNN)

• Category: Computer Vision - Image Classification
• Description: Implements a Convolutional Neural Network (CNN) to classify images from the CIFAR-10 dataset into 10 categories (e.g., planes, cars, birds, etc.).
• Features:
  • Two convolutional layers with ReLU activations and max-pooling.
  • Fully connected layers for classification.
  • Integrates GPU support for faster training.
  • Visualizes training and validation metrics.
• Notebook: cs480_winter23_asst3_cnn_cifar10.ipynb

5. Generation

• Category: Natural Language Processing (NLP) - Sequence Generation
• Description: Implements a character-level RNN to generate sequences, such as names, based on a given category (e.g., language of origin). The model predicts the next character iteratively.
• Features:
  • Generates realistic names from different languages.
  • Pre-trained on datasets of names categorized by language.
  • Implements dropout to add variety in generation.
• Notebook: Generation.ipynb

---

6. Translation

• Category: Natural Language Processing (NLP) - Machine Translation
• Description: Implements a Sequence-to-Sequence (Seq2Seq) model with an attention mechanism to translate text from French to English.
• Features:
  • Encoder-Decoder architecture with attention.
  • Handles paired datasets for machine translation tasks.
  • Demonstrates translation with variable-length sequences.
• Notebook: Translation.ipynb

---

7. Classification

• Category: Natural Language Processing (NLP) - Text Classification
• Description: Uses a character-level RNN to classify words into categories (e.g., predicting the language of origin for a given name).
• Features:
  • Processes variable-length input sequences.
  • Predicts the language of origin based on the spelling of names.
  • Includes training and evaluation on categorized datasets.
• Notebook: Classification.ipynb

---

8. Variational Autoencoder (VAE)

• Category: Generative Modeling
• Description: Implements a Variational Autoencoder (VAE) for generating new samples based on input data by learning a latent representation.
• Features:
  • Encodes data into a latent space and decodes it to reconstruct data with added variability.
  • Includes reparameterization trick for gradient-based optimization.
  • Trained on the MNIST dataset for image generation.
• Notebook: VAE_Complete.ipynb

---

9. Generative Adversarial Network (GAN)

• Category: Generative Modeling
• Description: Implements a Generative Adversarial Network (GAN) consisting of a generator and discriminator that compete to produce realistic images.
• Features:
  • Generator network creates synthetic images.
  • Discriminator network evaluates the authenticity of images.
  • Trained on the MNIST dataset to generate realistic handwritten digits.
• Notebook: GAN Complete.ipynb

Setup Instructions

Requirements

• Python 3.7+
• Libraries:
  • numpy
  • matplotlib
  • tqdm (for Logistic Regression)
  • torch
  • torchvision

Installation

1. Install dependencies:

    pip install numpy matplotlib tqdm torch torchvision`

Running the Notebooks

1. Clone this repository:

   git clone https://github.com/SV592/ML.git
   cd ML