PSI - Issue 70

Available online at www.sciencedirect.com

ScienceDirect

Procedia Structural Integrity 70 (2025) 394–400

Keywords: MachineLearningML;GeopolymerCompositesGPC;ArtificialIntelligenceAI; Compressive Strength CS; Random Forest RF; Decision Tree DT; Abstract The manufacture and use of ordinary Portland cement (OPC) contribute to environmental harm and pose risks to human health due to natural resource depletion and greenhouse gas emissions. Efforts are currently focused on finding substitutes for OPC through the exploration of alternative binders. Aluminosilicate waste could potentially be utilised to create geopolymer. The research on geopolymer concrete (GPC) is expanding. Efforts and finances are invested in creating samples, treatments, and evaluations. Swift and economical research necessitates creative approaches. This study employed supervised machine learning (SML) techniques to forecast GPC compressive strength (CS) utilising decision tree (DT) and random forest (RF) algorithms. All models were validated and compared using R2 and statistical analysis. Solo SML methods demonstrated lower accuracy in predicting GPC CS compared to ensembles. Furthermore, the results of each individual SML model were commendable. The R2 values for the RF and DT models were 0.95 and 0.88, respectively. The lower MAE and RMSE values of the ensemble SML models confirmed their accuracy. RF outperforms DT. SML methods enable builders to evaluate material quality efficiently and cost-effectively. © 2025 The Authors. Published by ELSEVIER B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under the responsibility of International Conference on Structural Integrity Organizers 1. Introduction Cement concrete is the most commonly used construction material worldwide. It is composed of aggregates, water, and binding agents like Ordinary Portland Cement (OPC) (Khan et al., 2020; Khan & Ali, 2018; Khan et al., 2021). OPC accounts for 7% of the energy consumed by the industry, ranking it third after aluminium and steel (Teja et al., 2017; Gopalakrishnan & Kaveri, 2021; Yang et al., 2022). Cement production alone contributes nearly 8% of global Structural Integrity and Interactions of Materials in Civil Engineering Structures (SIIMCES-2025) Prediction of Geopolymer Composite Strength Using Soft Computing Techniques Sahil Sehrawat a , Ritesh Kumar Roushan b, *, Priyanka Rani b a Department of Computer Science and Engineering, Faculty of Engineering & Technology, SRM University, Sonepat, Haryana, India b Department of Civil Engineering, Faculty of Engineering & Technology, SRM University, Sonepat, Haryana, India

* Corresponding author. Tel.: +91-7549999012. E-mail address: riteshroushan440@gmail.com

2452-3216 © 2025 The Authors. Published by ELSEVIER B.V. This is an open access article under the CC BY-NC-ND license (https://creativecommons.org/licenses/by-nc-nd/4.0) Peer-review under the responsibility of International Conference on Structural Integrity Organizers 10.1016/j.prostr.2025.07.069