Skip to content

Latest commit

 

History

History
29 lines (21 loc) · 1.15 KB

README.md

File metadata and controls

29 lines (21 loc) · 1.15 KB

Powerpredict

Daily power consumption is related to the weather (rain, sunshine, temperature, etc). Prediction of the power consumption based on the weather is relevant for energy suppliers. This project is about creating a Machine Learning Model to predict the power consumption according to given data.

The Dataset:

In the given Dataset we have information about temperature, humidity, etc. for the five citys:

  • Bedrock
  • Gotham City
  • New New York
  • Paperopoli
  • Springfield

and one column that shows the power consumption.

Scores for different models:

1. Running on the 20 features with highest correlation to power consumption

Name Training Set Score Test Set Score Hidden Score
Linear Regression 3422.96 3403.61 ?
Polynomial Linear Regession 3152.17 3162.33 3320.30
K Nearest Neighbors Regrssor 954.26 1881.62 2503.17
Random Forest Regressor 679.13 2314.29 3477.68