Application of Machine Learning for Predicting Concrete Strength: Ensembles vs. Instance-Based Algorithms in WEKA

Abstract

This research work presents a comprehensive analysis of machine learning (ML) techniques for predicting the compressive strength of concrete, a critical parameter in civil engineering. The study compares instance-based learning methods, such as Locally Weighted Learning (LWL), K*, and IBk, with ensemble-based methods like Bagging, Random Committee, and Ensemble Selection, using the WEKA software platform. The research highlights the advantages of each ML approach, with ensemble methods generally outperforming instance-based methods in terms of prediction accuracy. The document also discusses the importance of data preprocessing, particularly the handling of outliers and extreme values, and employs Spearman's rank correlation for statistical analysis. The findings contribute to the advancement of ML applications in the construction industry, offering insights into the comparative strengths of different ML algorithms for predicting concrete compressive strength.