- Clone the repository
- If you want to use a virtual environment, create one and activate it
- Run
make setup
To make sure that everything is working correctly, you can run the tests by running make test.
If you want to run the code in the jupyter notebooks, make sure to install your virtual environment as a kernel
by running ipython kernel install --user --name=<your-virtual-env>
This code has been developed and tested with Python 3.12
The EDA.ipynb notebook performs an Exploratory Data Analysis (EDA) on a dataset. It analyzes
the different variable distributions, their missing values and the correlation between them.
You can open the notebook by running make eda.
The Model_Selection.ipynb notebook explores different data preprocessing techniques such as different
imputations methods, encoding, feature engineering and feature selection. It also explores different regression
models and finds that the best model is a Gradient Boosting Regressor with engineered features.
You can open the notebook by running make model-selection.
The Optimal_Pricing.ipynb notebook uses a simple Logistic Regression model to estimate the
customer demand and its response to price changes. The notebook uses this model to compute the optimal price
and explore different pricing strategies. For two different strategies, the expected revenue and demand,
and their volatilities are computed.
You can open the notebook by running make optimization.
The optimal pricing strategy is obtained by using the standard optimal price equation which relates marginal cost to marginal revenue. The Logistic Regression makes it simple to compute demand elasticity.
An alternative pricing strategy is explored where the optimization is performed given a constraint on the resulting demand, which should be no less than a given threshold.
The Bootstrap method is used to verify the expected confidence intervals for the revenue and demand estimates.
