Time Series Forecasting For Energy Consumption Using XGBoost and LSTM

The increasing demand for accurate energy consumption forecasting has spurred a surge of interest in leveraging advanced machine learning techniques. This project focuses on the application in XG Boost, a powerful and versatile gradient-boosting algorithm, time series forecasting (TSF) in context of energy consumption. LSTM networks are applied for forecasting of predicted energy consumption. We analyze the abilities of LSTM models in predicting patterns of consumption, which is very crucial for structured resource management and sustainability initiatives and it exhibits a performance of 82% aacuracy, Intending to improve the prediction accuracy and address the challenges posed by the dynamic nature of energy consumption patterns, we employ XG Boost to model intricate dependencies and capture non-linear relationships within the series data and it shows a performance of 89.05% accuracy. XGBoost has oupperformed LSTM for TSF in energy consumption.