Skip to content

Latest commit

 

History

History
79 lines (58 loc) · 2.05 KB

README.md

File metadata and controls

79 lines (58 loc) · 2.05 KB
Raw

Hi, welcome to mnist_ML

This repository contains my Jupyter Notebook analysis of several machine learning models for classifying handwritten digits from the MNIST dataset. The analysis explores a variety of models ranging from t-SNE for dimensionality reduction to deep learning with ResNet-9.

Table of Contents

Overview

This project is focused on comparing different machine learning models to classify the MNIST dataset, a well-known benchmark for handwritten digit classification. The models analyzed include:

  • t-SNE (t-Distributed Stochastic Neighbor Embedding) for dimensionality reduction
  • Traditional Machine Learning models such as Random Forest
  • Deep learning models including ResNet-9

The results are visualized through various plots to assess the performance of each model and compare their effectiveness in classifying the digits.

Requirements

To run the notebook, you'll need the following libraries installed:

matplotlib
numpy
pandas
seaborn
tensorflow
openTSNE
scipy
scikit-learn

Installation

  1. Clone the repository:

    git clone https://github.com/kyracho/mnist-classification.git
cd mnist-classification

  2. Set up a virtual environment (optional):

    python -m venv venv
source venv/bin/activate  # On Windows use `venv\Scripts\activate`

  3. Install the required packages:

    pip install -r requirements.txt

Alternatively, install the packages manually:

pip install matplotlib numpy pandas seaborn tensorflow openTSNE scipy scikit-learn

Running the Notebook

  1. Launch Jupyter Notebook:

    jupyter notebook

  2. Open the mnist_classification.ipynb file from the Jupyter interface.

  3. Run the cells in the notebook to train and evaluate various machine learning models.