Precision Prediction of Household Electricity Consumption Through Data-Driven Model

An effective strategy for managing energy and sustainability is the accurate forecasting of household electricity consumption. A new challenge arises in consumption patterns for traditional models, which face difficulties in variability and data variety. This study aims to bridge the gap by proposing a novel technique called the Mountain Gazelle optimizer-driven Malleable Random Forest technique (MG-MRF), for improving electricity consumption prediction. This has enabled MG-MRF to model different consumption patterns as well as manage variability in the data. The study collected extensive datasets from different households, and those datasets had to undergo preprocessing to ensure integrity. Evaluation results of the approach further underscore the potential of MG-MRF to give accurate and dependable predictions, consequently allowing informed decision-making for the consumption of energy. The proposed method outperformed the traditional models with a prediction accuracy of 98.2%, precision of 94%, recall of 90%, and an f1-score of 92%. This study emphasizes the importance of adaptive modeling techniques in understanding and predicting household electricity usage, enabling the development of more effective energy management strategies. The experimental results advocate and contribute to sustainable energy practices by raising consumer awareness regarding their electrical consumption.