Getting Started #1
Description
Tips for Getting Started
Hey hackers,
Ok, so this may seem like a daunting challenge. We gave you all this data, but not much else to work with. Here's some info on how to approach the problem.
- First, plot some data.
- What does it look like? Industrial sites tend to have more spiky electricity usage due to their use of motors. Commercial sites tend to have more rounded electricity usage that aligns with typical working hours.
- Pick a few days throughout the year, do you notice any trends? Electricity usage often has seasonal trends. - Second, do some research on load forecasting.
- Load forecasting (and time-series forecasting in general) is a heavily researched topic. You'll be able to find good resources to guide you in this challenge, both for machine learning techniques and for relevant supporting data. Think about what drives electricity usage. - Third, find some external resources.
- You'll be able to find some external data sources and libraries to aid you in this challenge. Go out and get the data you need to improve your forecast.
Remember, there are two parts to this challenge. The better your forecast, the better you'll do. But don't forget about the usefulness of energy storage in reducing peak energy consumption!
