Forecasting Residential Electricity Demand Through Machine Learning and Model Synthesis
Presentation Date
4-10-2019
Degree Program
Economics, MS
Major Advisor Name
Kelly Chen
Type of Submission
Scholarly Poster
Abstract
This paper aims to develop a predictive model of residential electricity demand using techniques from statistical science, data analysis and econometrics. Residential energy intensity is investigated as a critical component of demand and evaluated as a predictor of energy use per household using a panel data set compiled from the US Energy Information Administration. Statistical and machine learning methods are combined using an umbrella of linear regression, and predictive accuracy is tested for in-sample and out-of-sample validity.
COinS