The load profile on the power system network might fluctuate greatly due to diverse user load behavior and increasing load demand. In order to maintain balance between load demand and supply, Electric utilities have relied heavily on load forecasting. Accurate short term electric load forecasting is crucial to assist balance between production and consumption of electrical energy. This paper presents Implementation of hybrid machine learning based regression methodology for short-term load forecasting with reference to weather and time aspects. Ensemble learning technique-based stacking regressor which predicts combining multiple regressor models to improve predictive performance. It trains final estimator based on predictions of many base regressors. In this paper Ridge regressor, Decision tree regressor (DTR), Gradient Boost regressor (GBR), Random Forest regressor (RFR) are used as base regressor and Multiple Linear Regression (MLR) is considered as final-estimator. Model is tested on Talapady substation data considering weather information and load data. The implemented strategy exhibits robustness across different forecasting horizons, making it a promising solution for dynamic energy management. Further the research underscores the potential of advanced machine learning techniques in optimizing load forecasting processes, ultimately contributing to more resilient energy systems.