Machine learning intelligence to assess the shear capacity of corroded reinforced concrete beams

Kumar, Aman; Arora, Harish Chandra; Kapoor, Nishant Raj; Kumar, Krishna; Hadzima-Nyarko, Marijana; Radu, Dorin

doi:10.1038/s41598-023-30037-9

Cited by 14 publications

(4 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…SHAP (SHapley Additive exPlanations) is a unified approach to explain the output of any ML model. It uses Shapley values, a well-established mathematical concept from cooperative game theory, to explain the output of a model by assigning a contribution to each feature 64 .…”

Section: Resultsmentioning

confidence: 99%

Ensemble learning based compressive strength prediction of concrete structures through real-time non-destructive testing

Arora,

Bhushan,

Kumar

et al. 2024

Sci Rep

Self Cite

View full text Add to dashboard Cite

This study conducts an extensive comparative analysis of computational intelligence approaches aimed at predicting the compressive strength (CS) of concrete, utilizing two non-destructive testing (NDT) methods: the rebound hammer (RH) and the ultrasonic pulse velocity (UPV) test. In the ensemble learning approach, the six most popular algorithms (Adaboost, CatBoost, gradient boosting tree (GBT), random forest (RF), stacking, and extreme gradient boosting (XGB)) have been used to develop the prediction models of CS of concrete based on NDT. The ML models have been developed using a total of 721 samples, of which 111 were cast in the laboratory, 134 were obtained from in-situ testing, and the other samples were gathered from the literature. Among the three categories of analytical models—RH models, UPV models, and combined RH and UPV models; seven, ten, and thirteen models have been used respectively. AdaBoost, CatBoost, GBT, RF, Stacking, and XGB models have been used to improve the accuracy and dependability of the analytical models. The RH-M5, UPV-M6, and C-M6 (combined UPV and RH model) models were found with highest performance level amongst all the analytical models. The MAPE value of XGB was observed to be 84.37%, 83.24%, 77.33%, 59.46%, and 81.08% lower than AdaBoost, CatBoost, GBT, RF, and stacking, respectively. The performance of XGB model has been found best than other soft computing techniques and existing traditional predictive models.

show abstract

Section: Resultsmentioning

confidence: 99%

Ensemble learning based compressive strength prediction of concrete structures through real-time non-destructive testing

Arora,

Bhushan,

Kumar

et al. 2024

Sci Rep

Self Cite

View full text Add to dashboard Cite

show abstract

“…The most common data normalization range used in the prediction of models are -1 to 1, 0 to 1, or 0.1 to 0.9, etc. 42 . In this article, the data values were normalized in the range of -1 to 1 using Eq.…”

Section: Methodology Of Studymentioning

confidence: 99%

Prediction of axial capacity of corrosion-affected RC columns strengthened with inclusive FRP

Kumar,

Arora,

Kumar

et al. 2024

Sci Rep

Self Cite

View full text Add to dashboard Cite

The primary cause behind the degradation of reinforced concrete (RC) structures is the propagation of corrosion in the steel-RC structures. Nowadays, numerous retrofitting techniques are available in the construction sector. Fiber-reinforced polymer (FRP) is one of the efficient rehabilitation measures that can be implemented on corroded structures to enhance structural capacities. However, the estimation of axial strength of FRP-strengthened columns affected by corrosion has been a challenging and tedious task in the laboratory as well as on the site. Considering such shortcomings, the prediction of axial capacity can be done using various analytical methods and artificial intelligence (AI) techniques. In this study, a comprehensive dataset of circular columns was extracted from the literature to predict the axial strength of FRP-wrapped and unstrengthened RC corroded columns. The laboratory results from the assembled dataset were compared to corresponding values estimated using relevant design codes provided by American Concrete Institute (ACI 440.2R-17 and ACI 318-19), and Bureau of Indian Standard (IS 456:2000). Five machine learning models were employed on columns to predict the axial load carrying capacity of FRP-strengthened and un-strengthened RC corroded columns. The results discovered that the extreme gradient boosting (XGBoost) model achieves superior accuracy with the least errors and could be used by the scientific community and FRP applicators to forecast the axial performance of corroded columns strengthened with and without FRP. The findings from the design codes revealed that prediction errors were available in high margins. Furthermore, feature importance analysis was conducted using the Shapley Additive exPlanation algorithm to know the contribution and influence of each input parameter on axial capacity. The feature analysis found that unconfined compressive strength of concrete plays an important role in deciding the axial capacity of columns. Moreover, to enhance the precision of axial capacity computation and improving the overall efficacy in engineering practice, a web-based user-friendly interface was developed for FRP applicators and engineers to simplify the process.

show abstract

“…The algorithm works by recursively partitioning the dataset into increasingly homogeneous subsets based on the values of input features [35]. Starting at the root node, which represents the entire dataset, the algorithm searches for the best feature to split the data, aiming to create subsets that are as pure as possible in terms of class distribution [36]. This process continues down the tree, with each internal node representing a decision point based on a specific feature.…”

Section: E K-nearest Neighbor (Knn) Classifiermentioning

confidence: 99%

Comparative Evaluation Of Machine Learning Classifiers For Brain Tumor Detection

Ali

2024

Preprint

View full text Add to dashboard Cite

Brain tumors, which are abnormal growths of cells in the brain, represent a significant health concern, necessitating prompt and accurate detection for effective treatment. If left untreated, brain tumors can lead to severe complications, including cognitive impairment, paralysis, and even death. This study evaluates six machine learning classifiers: Support Vector Classifier (SVC), Logistic Regression Classifier, K-Nearest Neighbors (KNN) Classifier, Naive Bayes Classifier, Decision Tree Classifier, and Random Forest Classifier - on a comprehensive brain tumor dataset. Our results showed that Random Forest achieved the highest accuracy of 98.27%, demonstrating its potential in detecting brain tumors. However, Support Vector Classifier (SVC) emerged as the top performer, achieving an impressive accuracy of 97.74%, showcasing its exceptional ability to detect brain tumors accurately. This significant improvement in SVC’s performance highlights its potential as a reliable tool for medical diagnostics, contributing to the development of efficient and accurate automated systems for early brain tumor diagnosis, ultimately aiming to improve patient outcomes and treatment efficacy.

show abstract

Machine learning intelligence to assess the shear capacity of corroded reinforced concrete beams

Cited by 14 publications

References 47 publications

Ensemble learning based compressive strength prediction of concrete structures through real-time non-destructive testing

Ensemble learning based compressive strength prediction of concrete structures through real-time non-destructive testing

Prediction of axial capacity of corrosion-affected RC columns strengthened with inclusive FRP

Comparative Evaluation Of Machine Learning Classifiers For Brain Tumor Detection

Contact Info

Product

Resources

About