Energy consumption - Time Series forecasting

The goal of this notebook is to provide some interesting insights into the consumption of a regional transmission network and macro consumption trends on a multi-state level.

All the available data is straightforward with no missing values and limited space for feature-engineering. As such, the focus will be on creating features based on the time-series column to provide additional information to the algorithm.

This was a fun and short project where I could experiment with different time-series options, lag functions and the rolling values method.

The complete notebook can be accessed here.

Dataset

The data is provided by www.pjm.com

• 2 features/columns
• 145,000 rows of data

Requirements

• Python 2.7 or Python 3.6
• Jupyter Notebook

License

MIT. See the LICENSE file for the copyright notice.

References:

1. https://arxiv.org/pdf/1603.02754v1.pdf
2. https://en.wikipedia.org/wiki/Regional_transmission_organization_(North_America)
3. https://www.pjm.com/about-pjm.aspx