https://github.com/hsm207/rossmann-fastai

My experiments on the rossmann dataset
https://github.com/hsm207/rossmann-fastai

deep-learning fastai pytorch tabular-data

Last synced: 8 months ago
JSON representation

My experiments on the rossmann dataset

• Host: GitHub
• URL: https://github.com/hsm207/rossmann-fastai
• Owner: hsm207
• License: gpl-3.0
• Created: 2019-06-01T13:22:26.000Z (over 6 years ago)
• Default Branch: master
• Last Pushed: 2019-06-01T22:04:44.000Z (over 6 years ago)
• Last Synced: 2025-01-14T11:16:27.849Z (9 months ago)
• Topics: deep-learning, fastai, pytorch, tabular-data
• Language: Jupyter Notebook
• Size: 218 KB
• Stars: 0
• Watchers: 3
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# Introduction

This repository contains my experiments on the [Rossmann Store Sales](https://www.kaggle.com/c/rossmann-store-sales/data) dataset

using the [fastai](https://github.com/fastai/fastai) library.

The goal is to investigate if there are any general techniques (e.g. preprocessing steps, feature engineering, etc)

that can help deep learning models on tabular data converge faster and/or with higher accuracy.

# Usage

1. Install Docker

2. Navigate to this project's root directory

3. Build and run this project's Dockerfile by executing:

    

    ```bash

    docker build -t fastai . && \

    docker run --runtime=nvidia \

               --name fastai \

               -p 8888:8888 \

               --dns 8.8.8.8 \

               fastai jupyter notebook --ip 0.0.0.0 --allow-root

    ```

      This will start a jupyter server on port 8888.

      

4. Open a web browser and navigate to the URL displayed in the previous step.

5. Navigate to the [notebooks](/notebooks) folder. This folder contains:

    * [01_rossmann_data_clean](/notebooks/01_rossman_data_clean.ipynb): Download the training and test set

    * [02_EDA](/notebooks/02_EDA.ipynb): Some exploratory data analysis on the dataset

    * [03_modelling](/notebooks/03_modelling.ipynb): Models/techniques I've tried so far

    

    # Contributing

    Feel free to raise a pull request for any questions, comments, feedback, etc.