1
$\begingroup$

I have a project to make a long term prediction (like 5 years) of electricity production by types of power plants (solor, wind, coal, nuclear etc.).

I have access to time series data in MW [megawatts] for days, weeks, months and years since 2010 until now.

What would you say are the best ML models for such task? Can you possibly refer me to some good resources regarding this topic you found helpful?

Thanks a lot.

Improve this question
$\endgroup$

1 Answer 1

Reset to default
0
$\begingroup$

I would probably do it by finding the best equation that fits your graph. If it's linear, I don't think you'll have many problems, it's just matter of performing a linear regression. Else, you can try applying a Bayesian information criterion to see which polynomial fits well on your data.

Remember to avoid overfitting: a 175-degree polynomial may fit optimally to your train dataset, but extremely poorly on future data. Usually, low grade polynomials are the best when performing this type of task.

Also, you could try assigning weights to your data: the production data of 2021 are surely more important than the ones from 2011.

Improve this answer
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.