K-fold Cross-validation Approach Research Articles

A comparative study on machine learning regression algorithms aplied to modeling gas centrifuge / Um estudo comparativo sobre algoritmos de regressão de aprendizagem de máquinas aplicado à modelagem de centrífugas a gás

The gas Centrifuge is a very hard equipment to model, because it involves a gas dynamic with many complications, such as hypersonic waves and rarefied regions combined with continuous flow areas. Therefore, data analysis regressions remain currently a very important technique to understand and describe the problem in a practical way. This paper intends to apply and compare several regression techniques using machine learning, to obtain a hydraulic and a separative power model of gas centrifuge used in enrichment plants. For this purpose, a set of normalized data composed of 134 experimental lines was used, observing the variables of interest, the separation power (dU), and the waste pressure (Pw), through the following explanatory variables: feed flow (F), cut (q), and product pressure (Pp). The comparisons were presented between the results obtained for the models generated by the following: algorithms, multivariate regression, multivariate adaptive regression splines – MARS, bootstrap aggregating multivariate adaptive regression splines – Bagging MARS, artificial neural network – ANN, extreme gradient boosting – XGBoost, support vector regression– Poly SVR, radial basis Function support vector regression – RBF SVR, K-nearest neighbors – KNN and Stacked Ensemble. That way, to avoid overfitting and provide insights about generalization of the models in unseen data, during the training phase, the k-fold cross validation approach was used. Subsequently, the residuals were analyzed, and the models were compared by the following metrics: Root mean square error – RMSE; Mean squared error – MSE; Mean absolute error – MAE; and Coefficient of determination – R2.

A Machine Vision Approach for Classification of Skin Cancer Using Hybrid Texture Features.

The main purpose of this study is to observe the importance of machine vision (MV) approach for the identification of five types of skin cancers, namely, actinic-keratosis, benign, solar-lentigo, malignant, and nevus. The 1000 (200 × 5) benchmark image datasets of skin cancers are collected from the International Skin Imaging Collaboration (ISIC). The acquired ISIC image datasets were transformed into texture feature dataset that was a combination of first-order histogram and gray level co-occurrence matrix (GLCM) features. For the skin cancer image, a total of 137,400 (229 × 3 x 200) texture features were acquired on three nonover-lapping regions of interest (ROIs). Principal component analysis (PCA) clustering approach was employed for reducing the dimension of feature dataset. Each image acquired twenty most discriminate features based on two different approaches of statistical features such as average correlation coefficient plus probability of error (ACC + POE) and Fisher (Fis). Furthermore, a correlation-based feature selection (CFS) approach was employed for feature reduction, and optimized 12 features were acquired. Furthermore, a classification algorithm naive bayes (NB), Bayes Net (BN), LMT Tree, and multilayer perception (MLP) using 10 K-fold cross-validation approach were employed on optimized feature datasets and the overall accuracy achieved by MLP is 97.1333%.

Ensemble hybrid machine learning methods for gully erosion susceptibility mapping: K-fold cross validation approach

Gully erosion is one of the important problems creating barrier to agricultural development. The present research used the radial basis function neural network (RBFnn) and its ensemble with random sub-space (RSS) and rotation forest (RTF) ensemble Meta classifiers for the spatial mapping of gully erosion susceptibility (GES) in Hinglo river basin. 120 gullies were marked and grouped into four-fold. A total of 23 factors including topographical, hydrological, lithological, and soil physio-chemical properties were effectively used. GES maps were built by RBFnn, RSS-RBFnn, and RTF-RBFnn models. The very high susceptibility zone of RBFnn, RTF-RBFnn and RSS-RBFnn models covered 6.75%, 6.72% and 6.57% in Fold-1, 6.21%, 6.10% and 6.09% in Fold-2, 6.26%, 6.13% and 6.05% in Fold-3 and 7%, 6.975% and 6.42% in Fold-4 of the basin. Receiver operating characteristics (ROC) curve and statistical techniques such as mean-absolute-error (MAE), root-mean-absolute-error (RMSE) and relative gully density area (R-index) methods were used for evaluating the GES maps. The results of the ROC, MAE, RMSE and R-index methods showed that the models of susceptibility to gully erosion have excellent predictive efficiency. The simulation results based on machine learning are satisfactory and outstanding and could be used to forecast the areas vulnerable to gully erosion.

Examining Techniques to Solving Imbalanced Datasets in Educational Data Mining Systems

The educational data mining research attempts have contributed in developing policies to improve student learning in different levels of educational institutions. One of the common challenges to building accurate classification and prediction systems is the imbalanced distribution of classes in the data collected. This study investigates data-level techniques and algorithm-level techniques. Six classifiers from each technique are used to explore their effectiveness to handle the imbalanced data problem while predicting students’ graduation grade based on their performance at the first stage. The classifiers are tested using the k-fold cross-validation approach before and after applying the data-level and algorithm-level techniques. For the purpose of evaluation, various evaluation metrics have been used such as accuracy, precision, recall, and f1-score. The results showed that the classifiers do not perform well with imbalanced dataset, and the performance could be improved by using these techniques. As for the level of improvement, it varies from one technique to another. Additionally, the results of the statistical hypothesis testing confirmed that there were no statistically significant differences for classifiers of the two techniques.

A statistical learning perspective on the inversion of NMR relaxation data

T 2 distribution is a powerful tool in the low-field nuclear magnetic resonance technique. The T2 distribution obtained from time-domain data involves an ill-posed inverse Laplace transformation. Tikhonov regularization with an L2 penalty term is most commonly used in this kind of problem, and the discrepancy principle, generalized cross-validation, L-curve, and S-curve methods are widely used in the selection of the regularization parameter. However, these selection approaches require prior knowledge, such as an accurate estimation of the threshold level of noise or setting a default value. In this paper, we propose a new method—the stability-enhanced k-fold cross-validation (SECV) approach—to perform a robust automatic search for the regularization parameter from a statistical learning perspective. In addition to considering test set residuals, additional terms—the Pearson’s correlation coefficients of the solutions of the disjoint subsets—are put forward to enhance the stability of the solution and make a trade-off between its imitative effect and interpretability. A bimodal T2 distribution model was constructed, and abundant echo trains with different noise levels were generated for the validation of the proposed method. The relative error of the estimates is used as a measure to evaluate the performance. The inversion results from the SECV method were compared with the solutions from the conventional methods, and the results showed that the proposed method is robust without manual intervention and suitable for both low- and high-signal-to-noise ratio data. Finally, mercury injection and nuclear magnetic resonance experiments were carried out on rock core samples to verify the correctness of our method.

Artificial neural network model for predicting the local compression capacity of stirrups-confined concrete

The local compression capacity is a key mechanical property for several constructions, such as the post-tensioned prestressed concrete. Due to the disordered stress distribution caused by the introduction of the concentrated load, it is a huge challenge to predict accurately the local compression capacity of stirrups-confined concrete to avoid the whole structure damage caused by local failure. This study explored a new approach to obtain the local compression capacity for stirrups-confined concrete using artificial neural networks (ANN). The ANN model was trained by a reliable database consisting of 180 samples from previous literature, which had a large application range, covering concrete strengths 15–112 MPa while the stirrup yield strengths 230–660 MPa. The main parameters, including the concrete strength, the local area aspect ratio, the core area aspect ratio, the ratio of duct diameter to the section width, the yield strength of stirrups, and the volumetric ratios of stirrups, were selected as input variables. To evaluate the proposed ANN model, the k-fold cross-validation approach was adopted to determine the generalization and reliability, and the sensitivity analysis was conducted to investigate the importance of the input parameters. Moreover, an empirical ANN equation was generated based on the proposed ANN model for design. Finally, the ANN model and the ANN equation were evaluated and verified against experimental data and existing models. The findings indicated that the ANN approach was highly applicable and reliable for estimating the local compression capacity of stirrups-confined concrete.

Estimating Automobile Crash Characteristics from Images using Deep Learning


 
 
 Crash characteristics such as crash velocity (Delta-V) and location of collision (LOC) are important determinants of the severity of the injury sustained by an occupant of an accident vehicle. Based on the predicted severity levels of injury, insurance companies can estimate the claim’s cost and better plan their financial reserves. We present a promising approach for accurately predicting Delta-V and LOC using deep learning methods, without the need for a forensic crash reconstruction. We constrain the study to small passenger vehicles and to front and rear collisions with crash velocities under 96 kph. We first develop and refine our image processing and deep CNN architectures using images created by using vehicle crash simulation software. Using a k-fold cross-validation approach, our methods are able to predict the crash velocity of simulated collisions (108 images) with a MAE of 3.41 kph (MAPE of 8.2%). Similarly, a multiple task learning CNN is able to predict Delta-V of real-world collisions (310 images) with a MAE of 4.19 kph (MAPE of 16.2%) and classify the LOC with 92% accuracy.
 
 

Vehicle Traffic Estimation Using Deep Learning

For commuters, vehicular traffic is an important planning concern. People have access to the weather forecast and the current traffic situation, but there is no application available to estimate traffic congestion and flow in the near future. Thus, we design and develop a machine learning approach which can predict vehicular traffic density and flowrate up to two days in the future based on the weather, calendar and special events data.First, Convolutional Neural Network (CNN) and Long Short Term Memory (LSTM) networks are utilized to predict the number of new vehicles and the total number of vehicles in images captured by a Nova Scotia Webcams (NS Webcams) video camera. The best models provide a Mean Absolute Percentage Error (MAPE) of 20.38% for the number of new vehicles and 18.56% for the total number of vehicles. These values are used to estimate traffic flowrate and density for hourly records over a three-month period.The hourly traffic data is combined with observed and forecasted weather data, and special event data to create a time series data. A Multiple Task Learning (MTL) - LSTM model is trained and tested using these data and a K-fold cross-validation approach. The Mean Absolute Error (MAE) and MAPE are used to evaluate the model performance. The MTL-LSTM model achieves a MAPE of 19.35% and 27.50% for flowrate and density using observed weather data, respectively. In the case of forecasted weather data, the MAPE for flowrate and density increases to 20.51% and 31.10%, respectively.

Additive and Interactive Genetically Contextual Effects of HbA1c on cg19693031 Methylation in Type 2 Diabetes.

Type 2 diabetes mellitus (T2D) has a complex genetic and environmental architecture that underlies its development and clinical presentation. Despite the identification of well over a hundred genetic variants and CpG sites that associate with T2D, a robust biosignature that could be used to prevent or forestall clinical disease has not been developed. Based on the premise that underlying genetic variation influences DNA methylation (DNAm) independently of or in combination with environmental exposures, we assessed the ability of local and distal gene x methylation (GxMeth) interactive effects to improve cg19693031 models for predicting T2D status in an African American cohort. Using genome-wide genetic data from 506 subjects, we identified a total of 1476 GxMeth terms associated with HbA1c values. The GxMeth SNPs map to biological pathways associated with the development and complications of T2D, with genetically contextual differences in methylation observed only in diabetic subjects for two GxMeth SNPs (rs2390998 AG vs. GG, p = 4.63 × 10−11, Δβ = 13%, effect size = 0.16 [95% CI = 0.05, 0.32]; rs1074390 AA vs. GG, p = 3.93 × 10−4, Δβ = 9%, effect size = 0.38 [95% CI = 0.12, 0.56]. Using a repeated stratified k-fold cross-validation approach, a series of balanced random forest classifiers with random under-sampling were built to evaluate the addition of GxMeth terms to cg19693031 models to discriminate between normoglycemic controls versus T2D subjects. The results were compared to those obtained from models incorporating only the covariates (age, sex and BMI) and the addition of cg19693031. We found a post-pruned classifier incorporating 10 GxMeth SNPs and cg19693031 adjusted for covariates predicted the T2D status, with the AUC, sensitivity, specificity and precision of the positive target class being 0.76, 0.81, 0.70 and 0.63, respectively. Comparatively, the AUC, sensitivity, specificity and precision using the covariates and cg19693031 were only 0.71, 0.74, 0.67 and 0.59, respectively. Collectively, we demonstrate correcting for genetic confounding of cg19693031 improves its ability to detect type 2 diabetes. We conclude that an integrated genetic–epigenetic approach could inform personalized medicine programming for more effective prevention and treatment of T2D.

Digital Image Forgery Detection Using Deep Learning Models

One of the challenges to image trust in digital and online apps, as well as on social media, is the current situation. Image forgery detection is a technique for detecting and locating fabricated components in a modified image. A sufficient amount of features is necessary for good image forgery detection, which can be achieved using a deep learning model that does not require human feature engineering other handcraft feature techniques. In this paper we used the GoogleNet deep learning model to extract picture features and the Random Forest machine learning technique to determine whether or not the image was fabricated. The proposed approach is implemented on the publicly available benchmark dataset MICC-F220 with k-fold cross validation approach to split the data set in to training and testing dataset and also compared with the state-of-the-art approaches

Comparative Study of Optimum Medical Diagnosis of Human Heart Disease Using Machine Learning Technique With and Without Sequential Feature Selection

Predicting heart disease is regarded as one of the most difficult challenges in the health-care profession. To predict cardiac disease, researchers employed a variety of algorithms including LDA, RF, GBC, DT, SVM, and KNN, as well as the feature selection algorithm sequential feature selection. For verification, the system employs the K-fold cross-validation approach. These six strategies were used to conduct the comparative study. The Dataset for Cleveland, Hungray, Switzerland, and Long Beach V, as well as the Dataset Heart Statlog Cleveland Hungary, were used to assess the models performance. For both Hungary, Switzerland & Long Beach V and Heart Statlog Cleveland Hungary Dataset, Random Forest Classifier sfs and Decision Tree Classifier sfs produced the highest and almost identical accuracy values (100%, 99.40% and 100%, 99.76% respectively). The findings were compared to previous research that focused on cardiac prediction. In the future, we hope to extend the model even further so that it may be used with various feature selection techniques; another possibility is to use a random forest classifier. The major goal of this study is to improve on previous work by developing a new and unique technique for creating the model, as well as to make the model relevant and easy to use in real-world situations.

Splitting and Length of Years for Improving Tree-Based Models to Predict Reference Crop Evapotranspiration in the Humid Regions of China

To improve the accuracy of estimating reference crop evapotranspiration for the efficient management of water resources and the optimal design of irrigation scheduling, the drawback of the traditional FAO-56 Penman–Monteith method requiring complete meteorological input variables needs to be overcome. This study evaluates the effects of using five data splitting strategies and three different time lengths of input datasets on predicting ET0. The random forest (RF) and extreme gradient boosting (XGB) models coupled with a K-fold cross-validation approach were applied to accomplish this objective. The results showed that the accuracy of the RF (R2 = 0.862, RMSE = 0.528, MAE = 0.383, NSE = 0.854) was overall better than that of XGB (R2 = 0.867, RMSE = 0.517, MAE = 0.377, NSE = 0.860) in different input parameters. Both the RF and XGB models with the combination of Tmax, Tmin, and Rs as inputs provided better accuracy on daily ET0 estimation than the corresponding models with other input combinations. Among all the data splitting strategies, S5 (with a 9:1 proportion) showed the optimal performance. Compared with the length of 30 years, the estimation accuracy of the 50-year length with limited data was reduced, while the length of meteorological data of 10 years improved the accuracy in southern China. Nevertheless, the performance of the 10-year data was the worst among the three time spans when considering the independent test. Therefore, to improve the daily ET0 predicting performance of the tree-based models in humid regions of China, the random forest model with datasets of 30 years and the 9:1 data splitting strategy is recommended.

Forecasting the Bearing Capacity of the Driven Piles Using Advanced Machine-Learning Techniques

Estimating the bearing capacity of piles is an essential point when seeking for safe and economic geotechnical structures. However, the traditional methods employed in this estimation are time-consuming and costly. The current study aims at elaborating a new alternative model for predicting the pile-bearing capacity based on eleven new advanced machine-learning methods in order to overcome these limitations. The modeling phase used a database of 100 samples collected from different countries. Additionally, eight relevant factors were selected in the input layer based on the literature recommendations. The optimal inputs were modeled using the machine-learning methods and their performance was assessed through six performance measures using a K-fold cross-validation approach. The comparative study proved the effectiveness of the DNN model, which displayed a higher performance in predicting the pile-bearing capacity. This elaborated model provided the optimal prediction, i.e., the closest to the experimental values, compared to the other models and formulae proposed by previous studies. Finally, a reliable and easy-to-use graphical interface was generated, namely “BeaCa2021”. This will be very helpful for researchers and civil engineers when estimating the pile-bearing capacity, with the advantage of saving time and money.

Application of Advanced Machine Learning Approaches to Predict the Compressive Strength of Concrete Containing Supplementary Cementitious Materials.

The casting and testing specimens for determining the mechanical properties of concrete is a time-consuming activity. This study employed supervised machine learning techniques, bagging, AdaBoost, gene expression programming, and decision tree to estimate the compressive strength of concrete containing supplementary cementitious materials (fly ash and blast furnace slag). The performance of the models was compared and assessed using the coefficient of determination (R2), mean absolute error, mean square error, and root mean square error. The performance of the model was further validated using the k-fold cross-validation approach. Compared to the other employed approaches, the bagging model was more effective in predicting results, with an R2 value of 0.92. A sensitivity analysis was also prepared to determine the level of contribution of each parameter utilized to run the models. The use of machine learning (ML) techniques to predict the mechanical properties of concrete will be beneficial to the field of civil engineering because it will save time, effort, and resources. The proposed techniques are efficient to forecast the strength properties of concrete containing supplementary cementitious materials (SCM) and pave the way towards the intelligent design of concrete elements and structures.

Evaluation of shape factor impact on discharge coefficient of side orifices using boost simulation model with extreme learning machine data-driven

ABSTRACT In this paper, for the first time, the impact of the shape factor on the discharge coefficient of side orifices is evaluated using the novel Extreme Learning Machine (ELM) model. In addition, the Monte Carlo simulations (MCs) are applied to assess the accuracy of the modelling. Furthermore, the validation is conducted by means of the k-fold cross-validation approach (with k = 5). In other words, the most optimized number of hidden neurons is chosen to be equal to 30 and the results of all activation functions of the extreme learning machine are examined and the sigmoid activation function is selected for simulating the discharge coefficient. Subsequently, two modelling combinati0ons are introduced using the input parameters and five different extreme learning machine models are also developed. The analysis of the modelling results exhibits that the model with the shape factor is more accurate. The superior model is a function of all input parameters reasonably estimating the discharge coefficient. For example, the values of R and MAPE for this model are estimated to be 0.990 and 0.223, respectively. The results of the superior model are also compared with the empirical equations and it was shown that this model has higher accuracy.

Exploring Sentinel-1 and Sentinel-2 diversity for flood inundation mapping using deep learning

Identification of flood water extent from satellite images has historically relied on either synthetic aperture radar (SAR) or multi-spectral (MS) imagery. MS sensors are limited to cloud free conditions, whereas SAR imagery is plagued by noise-like speckle. Prior studies that use combinations of MS and SAR data to overcome individual limitations of these sensors have not fully examined sensitivity of flood mapping performance to different combinations of SAR and MS derived spectral indices or band transformations in color space. This study explores the use of diverse bands of Sentinel 2 (S2) through well-established water indices and Sentinel 1 (S1) derived SAR imagery along with their combinations to assess their capability for generating accurate flood inundation maps. The robustness in performance of S-1 and S-2 band combinations was evaluated using 446 hand labeled flood inundation images spanning across 11 flood events from Sen1Floods11 dataset which are highly diverse in terms of land cover as well as location. A modified K-fold cross validation approach is used to evaluate the performance of 32 combinations of S1 and S2 bands using a fully connected deep convolutional neural network known as U-Net. Our results indicated that usage of elevation information has improved the capability of S1 imagery to produce more accurate flood inundation maps. Compared to a median F1 score of 0.62 when using only S1 bands, the combined use of S1 and elevation information led to an improved median F1 score of 0.73. Water extraction indices based on S2 bands have a statistically significant superior performance in comparison to S1. Among all the band combinations, HSV (Hue, Saturation, Value) transformation of S2 bands provides a median F1 score of 0.9, outperforming the commonly used water spectral indices owing to HSV’s transformation’s superior contrast distinguishing abilities. Additionally, U-Net algorithm was able to learn the relationship between raw S2 based water extraction indices and their corresponding raw S2 bands, but not of HSV owing to relatively complex computation involved in the latter. Results of the paper establishes important benchmarks for the extension of S1 and S2 data-based flood inundation mapping efforts over large spatial extents.

An enhanced CAD system based on machine Learning Algorithm for brain MRI classification

The development of Computer-Aided Design (CAD) and Computer-Aided Manufacturing (CAM) systems in the past decade has led to a remarkable advance in biomedical applications and devices. Particularly, CAM and CAD systems are employed in medical engineering, robotic surgery, clinical medicine, dentistry and other biomedical areas. Hence, the accuracy and precision of the CAD and CAM systems are extremely important for proper treatment. This work suggests a new CAD system for brain image classification by analyzing Magnetic Resonance Images (MRIs) of the brain. Firstly, we use the proposed Downsized Rank Kernel Partial Least Squares (DR-KPLS) as a feature extraction technique. Then, we perform the classification using Support Vector Machines (SVM) and we validate with a k-fold cross validation approach. Further, we utilize the Tabu search metaheuristic approach in order to determine the optimal parameter of the kernel function. The proposed algorithm is entitled DR-KPLS+SVM. The algorithm is tested on the OASIS MRI database. The proposed kernel-based classifier is found to be better performant than the existing methods.

Machine Learning With GA Optimization to Model the Agricultural Soil-Landscape of Germany: An Approach Involving Soil Functional Types With Their Multivariate Parameter Distributions Along the Depth Profile

Societal demands on soil functionality in agricultural soil-landscapes are confronted with yield losses and environmental impact. Soil functional information at national scale is required to address these challenges. On behalf of the well-known theory that soils and their site-specific characteristics are the product of the interaction of the soil-forming factors, pedometricians seek to model the soil-landscape relationship using machine learning. Following the rationale that similarity in soils is reflected by similarity in landscape characteristics, we defined soil functional types (SFTs) which were projected into space by machine learning. Each SFT is described by a multivariate soil parameter distribution along its depth profile. SFTs were derived by employing multivariate similarity analysis on the dataset of the Agricultural Soil Inventory. Soil profiles were compared on behalf of differing sets of soil properties considering the top 100 and 200 cm, respectively. Various depth weighting coefficients were tested to attribute topsoil properties higher importance. Support vector machine (SVM) models were then trained employing optimization with a distributed multiple-population hybrid Genetic algorithm for parameter tuning. Model training, tuning, and evaluation were implemented in a nested k-fold cross-validation approach to avoid overfitting. With regards to the SFTs, organic soils were differentiated from mineral soils of various particle size distributions being partly influenced by waterlogging and groundwater. Further SFTs reflect soils with a depth limitation within the top 100 cm and high stone content. Altogether, with SVM predictive model accuracies between 0.7 and 0.9, the agricultural soil-landscape of Germany was represented with eight SFTs. Soil functionality with regards to the soil’s capacity to store plant-available water and soil organic carbon is well characterized. Four additional soil functions are described to a certain extent. An extension of the approach to fully cover soil functions such as nutrient cycling, agricultural biomass production, filtering of contaminants, and soil as a habitat for soil biota is possible with the inclusion of additional soil properties. Altogether, the developed data product represents the 3D multivariate soil parameter space. Its agglomerated simplicity into a limited number of spatially allocated process units provides the basis to run agricultural process models at national scale (Germany).

Non-destructive method of biomass and nitrogen (N) level estimation in Stevia rebaudiana using various multispectral indices

Unmanned Aerial Vehicle (UAV) based remote sensing is one of the modern techniques for crop management, which has been used in this study for biomass and Nitrogen (N) level estimations for Stevia rebaudiana, a medicinal crop used as an alternative to sugar as a natural sweetener. Different levels of nitrogen treatments were given to S. rebaudiana and the crops were harvested for biomass estimation. Mica sense Altum multispectral sensor on board was used for acquiring the image data of the crop. The linear regression model was used to examine the best vegetation index using K-fold cross validation approach. Excess Green Index (ExG) was identified as best vegetation index for biomass estimation (R2 = 0.7; RMSE = 23.77 g/m2; nRMSE = 29.14%), whereas Enhanced Normalized Difference Vegetation Index (ENDVI) was found as best predictor for Nitrogen (N) level estimation (R2 = 0.9; RMSE = 1.75 g/m2; nRMSE = 14.59%).

Bayesian optimization algorithm based support vector regression analysis for estimation of shear capacity of FRP reinforced concrete members

The use of fiber-reinforced polymer (FRP) rebars in lieu of steel rebars has led to some deviations in the shear behavior of concrete members. Several methods have been proposed to forecast the shear capacity of such members. Nonetheless, there are differences in the methods of considering the various parameters affecting shear capacity, and some of them provide widely scattered and conservative results. This paper presents a hybrid of the Bayesian optimization algorithm (BOA) and support vector regression (SVR) as a novel modeling tool for the prediction of the shear capacity of FRP-reinforced members with no stirrups. For this purpose, a large dataset of simply supported beams and unidirectional slabs reinforced with FRP were utilized. The model performance was assessed using several statistical performance indicators and compared with the Japan Society of Civil Engineers (JSCE), British Institution of Structural Engineers (BISE), Canadian Standard Association (CSA), and American Concrete Institute (ACI) design codes and guidelines, as well as some other artificial intelligence (AI) models. For development of the model, all the hyperparameters, i.e., kernel function type, epsilon, box constraint, and kernel scale, were optimized using the BOA technique. The k-fold cross validation approach was utilized to avoid overfitting of the model. It was found that the mean, median, standard deviation, minimum, maximum, and interquartile range of the developed hybrid model predictions are very close to the experimental results. The predicted results overlap the experimental data with a coefficient of determination of 95.5%. The plot of relative deviations and residual plots are scattered around the zero reference line with low deviation, which indicates that the model is reliable and valid. The error terms (e.g., mean absolute error, root mean square error) obtained for all specimens were 4.85 and 11.03, which are very low values. The correlation coefficient (R) and fractional bias (FB) were found to be 0.977 and 0.0033, which are very close to 1 and 0, respectively, thus implying a reliable prediction. The comparative investigations with other codes and guidelines show that the hybrid BOA–SVR model predictions are more accurate and robust than those of the other models.