Title: Ensemble machine learning regression technique to select the type of concrete as radiation shielding material

Authors: Debabrata Datta; S. Seema; S. Suman Rajest; Biswaranjan Senapati; S. Silvia Priscila; Deepak K. Sinha

Addresses: Department of Information Technology, Heritage Institute of Technology Kolkata, West Bengal, India ' Department of Computer Science, The Tamil Nadu Dr.Ambedkar Law University, Chennai, Tamil Nadu, India ' Department of Research and Development (R&D) and International Student Affairs (ISA), Dhaanish Ahmed College of Engineering, Chennai, Tamil Nadu, India ' Department of Computer and Information Science, Parker Hannifin Corp., IL, 60503, USA ' Department of Computer Science, Bharath Institute of Higher Education and Research, Chennai, Tamil Nadu, India ' School of Computer Science and Engineering, Jain University, Bengaluru, Karnataka, India

Abstract: The selection of exact material for shielding analysis is challenging in radiation protection. The primary objective of shielding analysis is to reduce radiation exposure to the occupational worker at their workplace. Generally, high-density concrete is selected as the shielding material to prevent accidental exposure to gamma and neutron radiation. Composite material or multilayer shielding materials are generally used to optimise the cost of concrete with maximum benefit to the society of occupational radiation workers. A surrogate model for concrete's overall strength using cement, fly ash, and coarse and fine aggregates is created using machine learning and ensemble learning. Ensemble learning in machine learning solves underfitting and overfitting problems when fitting a regression model for shielding analysis. As density increases, concrete overall strength decreases. Several samples of various types of concrete (different compositions) are collected as input data. Finally, a multi-attribute decision-making method is applied to select the appropriate type of concrete. The research presents the ensemble learning based regression technique coupled with multi attribute decision making method to recommend the exact variety of concrete for shielding gamma and neutron radiation.

Keywords: gamma and neutron; technique of order preference for similarity ideal solution; TOPSIS; type of concrete; radiation shielding material; ensemble machine learning; regression technique; mean square error; MSE; root mean square error; RMSE; landscape of materials science.

DOI: 10.1504/IJCIS.2025.148322

International Journal of Critical Infrastructures, 2025 Vol.21 No.4, pp.317 - 337

Received: 08 Oct 2023
Accepted: 12 Jan 2024
Published online: 02 Sep 2025 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article