Electrical energy is one of the indicators that express the welfare and modernization level of countries. It is a type of energy that needs to be consumed as soon as it is produced and is cumbersome to store. For this reason, the estimation accuracy of the consumption demand is important in order to meet the supply. In this study, Artificial Neural Networks (ANN), which are frequently used in applications, and Differential Polynomial Neural Networks (D-PNN), a new type of neural network, are compared in the electricity consumption estimating problem. While comparing the methods, the independent variable data of exports, imports, population, installed power, and gross domestic product were used as the inputs of the models, and the electricity consumption values of Türkiye in a certain time interval were estimated. As a result of the comparisons, it was seen that the D-PNN method gave good results in performance criteria, ranging from 52.5% to 58.8%, compared to ANN.