<![CDATA[In this era of globalization, electricity is one of the essential needs in human life. Electricity load prediction plays a crucial role in designing supply-demand operations to avoid losses from various aspects. In this study, short-term electricity load prediction is conducted per 30 minutes, aiming to achieve minimum prediction errors. Support Vector Regression (SVR) is used as the machine learning method for data classification, and fractal method is employed for dimension calculation and feature extraction from historical data. The results of this research are as follows: In the first experiment, short-term prediction was conducted without using the fractal method, resulting in a Mean Absolute Percentage Error (MAPE) of 2.85%. In the second experiment, short-term prediction was performed using the dataset that had undergone fractal calculation and feature extraction, leading to a lower MAPE of 2.32%. The prediction results using the fractal method obtained a lower MAPE compared to the non-fractal approach. Fractal significantly influences the calculation of short-term electricity load prediction using Support Vector Regression (SVR).]]>
