6thWave: AI News Hub

This article delves into the descriptive statistics and machine learning models applied to analyze and predict energy consumption in educational buildings. Data from 352 schools was meticulously collected, revealing insights into structural and operational characteristics. Key statistics include an average of 3.26 floors, 2,596.59 square meters of built area, and 410,968 kWh/year in energy consumption. Variability was observed, with standard deviations highlighting differences in school sizes and operational parameters. Data visualization through scatter matrices and parallel coordinate plots identified correlations, especially between total built area, roof area, AC capacity, and energy use. The Pearson correlation analysis further confirmed these relationships, emphasizing the impact of school size and AC capacity on energy consumption. Four machine learning models—Decision Tree, K-Nearest Neighbors (KNN), Gradient Boosting, and Long-Short-Term Memory (LSTM)—were developed to predict energy use. Gradient Boosting and LSTM emerged as the most effective, demonstrating superior accuracy and generalization. This comprehensive analysis underscores the importance of understanding data characteristics and selecting appropriate predictive models to enhance energy efficiency in schools.

Source.