Search

The precise prediction of energy consumption is crucial for businesses, companies, and households especially when it comes to planning energy purchases. An underestimated or overestimated forecast value may result in the use of energy inefficiently. The companies will face financial consequences for inefficient energy usage because energy production requires high costs. In this research, an energy consumption prediction model is proposed employing Light Gradient-Boosting Machine (ECP_LightGBM) and explainable artificial intelligence (XAI), respectively for forecasting. A household dataset is used in this study for the evaluation of our model and also to compare the results with previously published approaches. According to the results, our model achieved the lowest root mean square error. Furthermore, the interpretability investigation using XAI indicated that the feature name sub_metering_3 had a very strong impact on the model’s output which shows the consumption of energy by air conditioner and water heater. Lastly, this study can be helpful for household practitioners, offering the LightGBM model for precise energy prediction and giving guidance to leaders and policymakers, so they can allocate investments and energy resources more intelligently.