Review of the state of the art of machine models for household consumption prediction

Forecasting energy usage is a challenge that enables power suppliers to address particular behaviors. These activities that power suppliers may perform include finding out the customers' behavior in order to adapt their prices to their consumption or the intervals at which energy demand will be higher and have planned the adjustment of supply chains. To this end, an evaluation should be carried out of the methods that make it possible to predict the energy consumption of the future according to the consumption history and other parameters of the users themselves. In this paper we discuss the main machine-learning methods for the prediction of power consumption using a one-year data set of a shoe store. The revision made it possible to notice that for the data set applying Linear Regression and Support Vector Regression a success of 85.7% has been achieved with the best results provided