This repository provides a solution for a practice problem on SIGNATE: Boxed Lunch Sales Forecasting. The sales of boxed lunch at a certain company O in Chiyoda-ku are possibly related to many factors such as the menu, the weather, or the date. Our goal is to forecast the sales of the boxed lunch.
Dependency
- Python 3.9.2
- Numpy 1.19.2
- Pandas 1.2.4
- Scikit-learn 0.24.1
- Feature-engine 1.0.2
- Sklearn-pandas 2.1.0
- Matplotlib 3.4.1
- Seaborn 0.11.1
To perform exploratory data analysis and check the results, run code in notebooks/exploratory_data_analysis.
To perform machine learning modeling, you can execute
-
Jupyter Notebook code in notebooks/modeling, or
-
Python code in train_valid_models.py.