Load Forecasting

Dataset from UCI contains the  electricity consumption (in kWh) of 320 customers every 15 minutes from 2011 to 2014.     For  some customers, their electricity consumption data was recorded starting from a different time,  and the data before the first record was marked as 0.     As training model requires a starting point for the time series, and different customers  have different starting times, some features remain 0 before a certain time point.     Using all  the data for training would inevitably have a negative impact.     Therefore, the data  from 2011 is removed  and set the starti

Categories:
354 Views

Leaderboard (numbers are kW MAE):

Teams with more than 5 missing submissions are eliminated from the leaderboard.

 

Last Updated On: 
Wed, 11/17/2021 - 21:02

Accurate short-term load forecasting (STLF) plays an increasingly important role in reliable and economical power system operations. This dataset contains The University of Texas at Dallas (UTD) campus load data with 13 buildings, together with 20 weather and calendar features. The dataset spans from 01/01/2014 to 12/31/2015 with an hourly resolution. The dataset is beneficial to various research such as STLF.

Categories:
3312 Views