<p>The present study refers to the process of estimating the amount of electricity that customers would use over a given period. Short-term power consumption forecasting for single private residencies is one very important and simultaneously challenging topic. The drivers of electricity consumption, along with factors of power consumption, are such important things that, if known, go a long way in the development of effective strategies towards improvement in energy efficiency and reduction of carbon footprints. It also involves various other benefits associated with cost savings, environmental benefits and efficient resource management. The primary objective of this work is to improve the accuracy in the estimation of power consumption by availing the various capabilities of each optimizer in fine-tuning the parameters of the elastic net model for better-handling multicollinearity and feature selection. Further, compare the hybrid model results with the traditional estimating approaches to show improvement in the level of accuracy and efficiency. In this regard, three of the most novel nature-inspired optimization techniques, namely Giant Armadillo Optimization, Giant Trevally Optimizer and Black Widow Optimizer are employed in the development of the elastic net model for this study. Several prediction models are presented and their accuracy in estimating power consumption is assessed. Through comparisons using the root mean square error (RMSE) as the key assessment metric, several different methods are compared to identify the best model. In conclusion, it will be demonstrated that EN+GTO has the best RMSE, with a value of 1080.48 for the total data.</p>