Analysis of a city's electricity consumption data with machine learning methods

Electricity distribution companies are obliged to inform their subscribers about their electricity consumption through the "Market Management System". Electricity distribution companies inform their consumers who do not have consumption values at the end of the month, that is, they do not have reading data, by estimating consumption values. The aim of this thesis is to make estimations with machine learning methods as well as the consumption forecasting methodology determined by the Energy Market Regulatory Authority and which is still being used. In this study, analyses were performed using machine learning methods, linear regression and long-short-term memory (LSTM) methods in MATLAB environment. Root Mean Square Error (RMSE) method was determined as the success criterion and linear regression analysis method provided the most successful result of the methods used. In this context, the studies of linear regression long short term memory and Energy Market Regulatory Authority's methodology, a comparison of estimation methods is made, and observed that consistent and successful results are obtained.