This repository has the code and description of the modelling approach that I used for the Load track of the Global Energy Forecasting Competition: 2014, hosted on CrowdAnalytix.
final_report.md
: A summary of my modelling approach as well as a log of what changed week by weekgef2014_functions.R
: Various functions used to load data, derive variables, and build modelstrain_pca_model.R
: Extracting principal components from the weather variablestrain_gbm_models.R
: Training the quantile regression gradient boosting machine modelspredict_distribution.R
: Simulating weather and predicting the distribution on the next month