A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Applying Optimized Machine Learning Models for Predicting Unconfined Compressive Strength in Fine-Grained Soil
https://orcid.org/http://orcid.org/0000-0003-1361-0126
This research investigates the potential of optimized machine learning (OML) models for the prediction of UCS in fine-grained soil. Leveraging OML models, the study makes use of an extensive geotechnical dataset to estimate the UCS of soil extracted from the Fewa area situated in the Lesser Himalayan region. OML models were developed using the adaptive neuro-fuzzy inference system (ANFIS) coupled with six optimization algorithms: waterwheel plant algorithm (WPA), moth-flame optimization algorithm (MFOA), grasshopper optimization algorithm (GOA), wild horse optimizer (WHO), ant-bee colony (ABC), and firefly algorithm (FA). Among the models tested, the ANFIS-GOA model exhibited superior predictive ability, with R2 = 0.979 and R2 = 0.931 during the training and testing stages, respectively. Furthermore, the accuracy of the OML model developed has been confirmed through the evaluation of various workability parameters, order analysis, and Taylor’s plot and the implementation of an innovative approach called external validation. This comprehensive analysis concludes that the OML model offers a novel tool for geotechnical engineers to estimate the UCS of soil, thereby holding substantial implications in the field of foundation engineering. Also, considering the uniform geographical conditions prevalent in a significant portion of the Lesser Himalayan region, the study’s conclusions can be extrapolated to settings with similar soil conditions.
Applying Optimized Machine Learning Models for Predicting Unconfined Compressive Strength in Fine-Grained Soil
https://orcid.org/http://orcid.org/0000-0003-1361-0126
This research investigates the potential of optimized machine learning (OML) models for the prediction of UCS in fine-grained soil. Leveraging OML models, the study makes use of an extensive geotechnical dataset to estimate the UCS of soil extracted from the Fewa area situated in the Lesser Himalayan region. OML models were developed using the adaptive neuro-fuzzy inference system (ANFIS) coupled with six optimization algorithms: waterwheel plant algorithm (WPA), moth-flame optimization algorithm (MFOA), grasshopper optimization algorithm (GOA), wild horse optimizer (WHO), ant-bee colony (ABC), and firefly algorithm (FA). Among the models tested, the ANFIS-GOA model exhibited superior predictive ability, with R2 = 0.979 and R2 = 0.931 during the training and testing stages, respectively. Furthermore, the accuracy of the OML model developed has been confirmed through the evaluation of various workability parameters, order analysis, and Taylor’s plot and the implementation of an innovative approach called external validation. This comprehensive analysis concludes that the OML model offers a novel tool for geotechnical engineers to estimate the UCS of soil, thereby holding substantial implications in the field of foundation engineering. Also, considering the uniform geographical conditions prevalent in a significant portion of the Lesser Himalayan region, the study’s conclusions can be extrapolated to settings with similar soil conditions.
Applying Optimized Machine Learning Models for Predicting Unconfined Compressive Strength in Fine-Grained Soil
https://orcid.org/http://orcid.org/0000-0003-1361-0126
This research investigates the potential of optimized machine learning (OML) models for the prediction of UCS in fine-grained soil. Leveraging OML models, the study makes use of an extensive geotechnical dataset to estimate the UCS of soil extracted from the Fewa area situated in the Lesser Himalayan region. OML models were developed using the adaptive neuro-fuzzy inference system (ANFIS) coupled with six optimization algorithms: waterwheel plant algorithm (WPA), moth-flame optimization algorithm (MFOA), grasshopper optimization algorithm (GOA), wild horse optimizer (WHO), ant-bee colony (ABC), and firefly algorithm (FA). Among the models tested, the ANFIS-GOA model exhibited superior predictive ability, with R2 = 0.979 and R2 = 0.931 during the training and testing stages, respectively. Furthermore, the accuracy of the OML model developed has been confirmed through the evaluation of various workability parameters, order analysis, and Taylor’s plot and the implementation of an innovative approach called external validation. This comprehensive analysis concludes that the OML model offers a novel tool for geotechnical engineers to estimate the UCS of soil, thereby holding substantial implications in the field of foundation engineering. Also, considering the uniform geographical conditions prevalent in a significant portion of the Lesser Himalayan region, the study’s conclusions can be extrapolated to settings with similar soil conditions.
Applying Optimized Machine Learning Models for Predicting Unconfined Compressive Strength in Fine-Grained Soil
Transp. Infrastruct. Geotech.
Thapa, Ishwor (author) / Ghani, Sufyan (author)
Transportation Infrastructure Geotechnology ; 11 ; 2235-2269
2024-08-01
35 pages
Article (Journal)
Electronic Resource
English