Computation of Compressive Strength of GGBS Mixed Concrete using Machine Learning 
Swati1, Jitendra Khatti2, Kamaldeep Singh Grover3
1Swati, M.Tech. Scholar, Department of Civil Engineering, Rajasthan Technical University, Kota (Rajasthan), India.
2Jitendra Khatti, PhD Fellow, Department of Civil Engineering, Rajasthan Technical University, Kota (Rajasthan), India.
3Kamaldeep Singh Grover, Professor, Department of Civil Engineering, Rajasthan Technical University, Kota (Rajasthan), India.
Manuscript received on November 20, 2021. | Revised Manuscript received on November 26, 2021. | Manuscript published on November 30, 2021. | PP: 241-250 | Volume-10 Issue-4, November 2021. | Retrieval Number: 100.1/ijrte.D66311110421 | DOI: 10.35940/ijrte.D6631.1110421
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Concrete is a composite material formed by cement, water, and aggregate. Concrete is an important material for any Civil Engineering project. Several concretes are produced as per the functional requirements using waste materials or by-products. Many researchers reported that these waste materials or by-products enhance the concrete properties, but the laboratory procedures for determining the concrete properties are time-consuming. Therefore, numerous researchers used statistical and artificial intelligence methods for predicting concrete properties. In the present research work, the compressive strength of GGBS mixed concrete is computed using AI technologies, namely Regression Analysis (RA), Support Vector Machine (SVM), Decision Tree (DT), and Artificial Neural Networks (ANNs). The cement content (CC), C/F ratio, w/c ratio, GGBS (in Kg & %), admixture, and age (days) are selected as input parameters to construct the RA, SVM, DT, ANNs models for computing the compressive strength of GGBS mixed concrete. The CS_MLR, Link_ CS_ SVM, 20LF_CS_DT, and GDM_CS_ANN models are identified as the best architectural AI models based on the performance of AI models. The performance of the best architectural AI models is compared to determine the optimum performance model. The correlation coefficient is computed for input and output variables. The compressive strength of GGBS mixed concrete is highly influenced by age (curing days). Comparing the performance of optimum performance AI models and models available in the literature study shows that the optimum performance AI model outperformed the published models.
Keywords: Compressive strength; GGBS, Support Vector Machine, Artificial Neural Networks