Machine Learning Based Estimation of Power Consumption through Spline Regression

Abstract: Considering the current scenario of steady depletion of primary sources of energy, it is extremely critical to manage the consumption of power. Ensuring low wastage of power involves prediction of power consumption, which can enable generation of power according to the demand. This process is true for both micro-scale estimation and generation, for particular sectors in an urban or rural area, to macro-estimate of power requirements in terms of overall power consumed nationally, that can be further divided into sectors such as commercial and residential power consumption. The present work outlines a machine learning based linear spline regression-based approach to prediction of energy consumption which is an improvement on the standard spline regression based approach employed currently.

Keywords: Power Consumption, Estimation, Machine Learning, Linear Regression, Spline Regression