alessiosanvito/PowerConsumptionPrediction-TimeSeriesAnalysis

Time series forecasting is crucial in a data-driven world, helping to predict future events and identify trends. This project analyzed electricity consumption in Morocco using ARIMA, UCM, and machine learning methods to find the most accurate model. The study aimed to model and forecast consumption every 10 minutes from Dec 1-30, 2017, for Tetouan. Time series forecasting is an increasingly important activity in an ever-more data-driven world, as it allows for predicting future events based on historical data and therefore making informed decisions in various business areas. Furthermore, it helps to identify trends and patterns in the data, which can be used to improve business processes and increase efficiency. This project was carried out to analyze the trend of electricity consumption and compare different forecasting methods using various algorithms such as ARIMA, UCM, and machine learning. Therefore, the aim of this study was to model and forecast the time series of electricity consumption in Morocco. The performance comparison of ARIMA, UCM, and machine learning methods was carried out to identify the model that predicted the consumption values every 10 minutes with the highest accuracy (obtaining clean residuals and good linear memory) in the period from December 1, 2017, to December 30, 2017, for the city of Tetouan.