Time series forecasting on power consumption pattern with CatBoost and regression model

There are two columns in the power_data.csv: timestamp and hourly power consumption.
The goal of the model is to predict the pattern of power consumption, e.g. weekly, monthly, and seasonal patterns.

The details of data analysis, model comparison and selection, model building and evaluation, conclusion and recommendation are provided in the notebook

The script of final model is saved in model.py
R-squared score is used to evaluate our regression models.