Abstract
Load forecasting is one of the most important aspects of power systems and Smart City concept. Forecasting a load of a power system with great accuracy not only for a few weeks but also for several years is important in terms of both technical and economic aspects. Nowadays, there exist many methods that are modeled to forecast the consumer's load. This paper will mainly focus on developing forecasting model basing on two Machine Learning (ML) techniques. The first is the branch of RNN method, which is named as Long-Short Term Memory (LSTM). The second one is Support Vector Machine. Both methods solve the problems related with prediction and show high accuracy and precision in time-series application.