Random datasets and Jupyter notebooks. Using SciKit-Learn and PyTorch.
This repository contains a collection of AI projects using various datasets and tools. I created this to practice building AI projects. Feel free to contribute and stuff. Look around, I have added some pretty graphs and diagrams for you. Each directory has its own README.md file too.
| Project title | Main tool | Description | Hyperlink |
|---|---|---|---|
| Blobs Classification | PyTorch | Simple blob classification using PyTorch. | blobs.ipynb |
| Circles Classification | PyTorch | Classifying circular data points with PyTorch. | circles.ipynb |
| Moons Classification | PyTorch | Identifying moon-shaped clusters in data. | moons.ipynb |
| Moons Overfit | PyTorch | Overfitting example with moon data. | moons-overfit.ipynb |
| Iris Dataset Model | SciKit-learn | Classic Iris dataset classification. | iris.ipynb |
| Wine Classification | SciKit-learn | Classify wine based on it's features. | wine.ipynb |
| 8x8 Digits Classification | SciKit-learn | 8 by 8 resolution images of handwritten digits. | digits.ipynb |
| Breast Cancer Diagnosis | SciKit-learn | Recognize cancer given numerical attributes. | cancer.ipynb |
| MNIST Classification MLP | PyTorch | Classifying MNIST digits with an MLP. | linear-mnist.ipynb |
To get started with messing up my code (or fixing it), clone the repository and install the necessary dependencies:
git clone https://github.com/nickkipshidze/random-ml-projects
cd random-ml-projects
pip install -r requirements.txtBy the way, you can just copy all that commands at once and paste it in the terminal, it will work.
Navigate to the projects directory and start the Jupyter Notebook server with this command:
jupyter notebookOBVIOUSLY. From there you can open Jupyter Notebook in your browser and look around in the repository, check out the .ipynb files. Or you can just ignore all that and use a VSCode extension instead. The choice is yours.
Contributions are welcome! Please open an issue or submit a pull request.
This project is licensed under the DONT TOUCH MY SHIT licese. See the LICENSE file for more details.
For any questions or feedback, please contact [email protected]. And no, I will not change the repository's license, do not bother messaging.
