Search

This research mainly focuses on investigating the role of the interactions between cash flows, oil price uncertainty, and geopolitical risk on corporate cash holdings across 42 countries. We also investigate whether such interactions may differ across firm and country characteristics. The main results show that the interactions between cash flows interacting with oil price uncertainty and geopolitical risk have a positive correlation with corporate cash holdings. Additionally, the nexus between the interactions and corporate cash holdings is more prominent for financially constrained firms, firms located in highly competitive industries, or Asia and Europe. The main policy implication based on our findings is that policymakers should exercise caution and take into account the effect of the uncertainties of oil price movement and geopolitical risk when designing policies related to corporate decision makings. [ABSTRACT FROM AUTHOR]

In recent years, the application of Deep Learning (DL) for damage detection in Structural Health Monitoring (SHM) using time-series data has garnered significant attention from the global scientific community. By incorporating numerous filters and units in their architecture, these models facilitate feature extraction making the processing of large time-series datasets in SHM more effective. However, optimizing of the number of filters and units is a challenge, often relying on scientists' experience or previous models without a systematic method for measurement or automatic selection. Therefore, this study proposes a novel framework to enhance the accuracy and efficiency of SHM for damage detection in bridge structures by using metaheuristic algorithms, Electric Eel Foraging Optimization (EEFO) algorithm, to optimize hyperparameters of DL model. This DL model combines the advantages of each traditional model. Specifically, 1D Convolutional Neural Network (1DCNN) is employed for features extraction, Gated Recurrent Units (GRU) for identifying long-term dependencies, and Residual Networks (ResNet) for avoiding vanishing gradient problem which often happens in DL model during training, referred to as 1DCNN-GRU-ResNet. The time-series dataset generated from Cua Rao bridge is used to demonstrate the effectiveness of the proposed method. 1DCNN-GRU-ResNet after optimization (1DCNN-GRU-ResNet-opt) achieves 91.6 %, significantly surpassing traditional 1DCNN (82.5 %), GRU (79.9 %), 1DCNN-GRU (85.7 %), and 1DCNN-GRU-ResNet (89.3 %) in the test set. The proposed 1DCNN-GRU-ResNet-opt approach demonstrates considerable potential in practical applications for SHM, offering high accuracy and efficiency. • Introducing a novel deep learning (DL) model, the 1DCNN-GRU-ResNet, for damage detection. • Utilizes Electric eel optimization for hyperparameter tuning in proposed DL model. • Exceeds both conventional and unoptimized DL model accuracies. • Demonstrates outstanding accuracy in time-series data for damage detection. • Validates practical applicability for real-world bridge health monitoring. [ABSTRACT FROM AUTHOR]

There has been increasing research on wireless body area networks (WBANs) due to their wide range of applications, rapidly developing personal equipment and healthcare support devices. This paper investigates a WBAN model that consists of many sensors. The sensors are distributed around the body and communicate with access points (APs). In order to improve the throughput of the system, an algorithm is applied to select the AP that has the best channel conditions to receive information from every sensor. Moreover, the authors developed the max–min power control algorithm to control the transmit power of sensors in the uplink and APs in the downlink depending on their channel condition. The proposed algorithm is investigated based on several parameters, such as the number of sensors, the number of APs, the length of uplink and downlink training in samples, and so on. According to simulation results, the sensor with max–min power control achieves higher throughput in all considered scenarios, leading to a significant improvement in the WBAN system throughput for both cases of the uplink and downlink. [ABSTRACT FROM AUTHOR]

Abstract: The removal of noise patterns in handwritten images requires careful processing. A noise pattern belongs to a class that we have either seen or not seen before. In the former case, the difficulty lies in the fact that some types of noise patterns look similar to certain characters or parts of characters. In the latter case, we do not know the class of noise in advance which excludes the possibility of using parametric learning methods. In order to address these difficulties, we formulate the noise removal and recognition as a single optimization problem, which can be solved by expectation maximization given that we have a recognition engine that is trained for clean images. We show that the processing time for a noisy input is higher than that of a clean input by a factor of two times the number of connected components of the input image in each iteration of the optimization process. Therefore, in order to speed up the convergence, we propose to use fuzzy inference systems in the initialization step of the optimization process. Fuzzy inference systems are based on linguistic rules that facilitate the definition of some common classes of noise patterns in handwritten images such as impulsive noise and background lines. We analyze the performance of our approach both in terms of recognition rate and speed. Our experimental results on a database of real-world handwritten images corroborate the effectiveness and feasibility of our approach in removing noise patterns and thus improving the recognition performance for noisy images. [Copyright &y& Elsevier]

Abstract: This paper presents a novel cascade ensemble classifier system for the recognition of handwritten digits. This new system aims at attaining a very high recognition rate and a very high reliability at the same time, in other words, achieving an excellent recognition performance of handwritten digits. The trade-offs among recognition, error, and rejection rates of the new recognition system are analyzed. Three solutions are proposed: (i) extracting more discriminative features to attain a high recognition rate, (ii) using ensemble classifiers to suppress the error rate and (iii) employing a novel cascade system to enhance the recognition rate and to reduce the rejection rate. Based on these strategies, seven sets of discriminative features and three sets of random hybrid features are extracted and used in the different layers of the cascade recognition system. The novel gating networks (GNs) are used to congregate the confidence values of three parallel artificial neural networks (ANNs) classifiers. The weights of the GNs are trained by the genetic algorithms (GAs) to achieve the overall optimal performance. Experiments conducted on the MNIST handwritten numeral database are shown with encouraging results: a high reliability of 99.96% with minimal rejection, or a 99.59% correct recognition rate without rejection in the last cascade layer. [Copyright &y& Elsevier]

In this paper, we consider the following nonlinear wave equation: (1)utt−B(∥ux∥2)uxx=f(x,t,u,ux,ut,∥ux∥2),x∈(0,1),0,(2)ux(0,t)−h0u(0,t)=u(1,t)=0,(3)u(x,0)=u˜0(x), ut(x,0)=u˜1(x), where B,f,u˜0,u˜1 are given functions. In Eq. (1), the nonlinear terms B(∥ux∥2),f(x,t,u,ux,ut,∥ux∥2) depending on an integral ∥ux∥2=∫lower limit 0, upper limit 1 |ux(x,t)|2 dx. In this paper, we associate with problem (1)–(3) a linear recursive scheme for which the existence of a local and unique solution is proved by using standard compactness argument. In case of B∈CN+1(R+),B⩾b0>0,B1∈CN(R+),B1⩾0,f∈CN+1([0,1]×R+×R3×R+) and f1∈CN([0,1]×R+×R3×R+) we obtain from the following equation: utt−[B(∥ux∥2)+ϵB1(∥ux∥2)]uxx=f(x,t,u,ux,ut,∥ux∥2)+ϵf1(x,t,u,ux,ut,∥ux∥2) associated to (2), (3) a weak solution uϵ(x,t) having an asymptotic expansion of order N+1 in ϵ, for ϵ sufficiently small. [Copyright &y& Elsevier]

• Nonlinear static stability problem of FG-TPMS shallow spherical shells is established. • Nine FG-TPMS designs by combining three TPMS structures and three porosity distributions are obtained. • Governing equations are derived using HSDT incorporating von Kármán's geometrical nonlinearities. • The Ritz energy minimization method is employed to achieve an efficient solution for the governing equations. • Buckling and postbuckling behaviors of FG-TPMS shallow spherical shells are analyzed numerically. An analytical solution for the nonlinear static stability problem of functionally graded triply periodic minimal surface (FG-TPMS) shallow spherical shells is studied in the current research for the first time. Three common TPMS structures including Primitive (P), Gyroid (G), and I-graph and Wrapped Package-graph (IWP) with three models of functionally graded porosity distribution along the thickness are considered. The shallow spherical shells (shallow SSs) are subjected to combined thermo-mechanical loadings and rested on a nonlinear elastic foundation. The fundamental formulas are expressed based on the higher-order shear deformation theory (HSDT) and von Kármán's geometrical nonlinearities. Employing the Ritz energy minimization method, the explicit relationship between load and deflection is derived. Subsequently, the static stability behavior of FG-TPMS shallow SSs is investigated. Numerical illustrations are investigated to show the superior thermo-mechanical load-carrying performance of the FG-TPMS SSs compared to corresponding isotropic structures of the same weight. The significant effects of geometric parameters, nonlinear elastic foundation parameters, and the type of FG-TPMS structures on the nonlinear static stability behavior of shallow SSs are further considered. [ABSTRACT FROM AUTHOR]

Epilepsy surgery is traditionally difficult to pursue in resource-limited countries but is nevertheless essential in the treatment of medication-refractory, surgically amenable epilepsy. With the help of international collaboration, a successful epilepsy program was started in Vietnam. This article comprises a retrospective chart review, combined with prospective longitudinal follow-up of 35 cases of unilateral drug-resistant epilepsy in the temporal lobe who underwent temporal lobectomy, in Viet Duc University Hospital from May 2018 to September 2022. The female/male ratio was 0.6:1, and focal seizures with impaired awareness accounted for 97.14% of patients. Of patients with focal awareness seizures, 51.41% were localized and detected by electroencephalography. Postoperatively, 80% of patients were seizure free (Engel I) at 1 year, and the remaining 20% had worthwhile seizure improvement (Engel II). Postoperative temporal lobe pathology was categorized as follows: mesial temporal sclerosis (48.57%), focal cortical dysplasia (25.71%), and low-grade neoplasms (25.71%). Of patients, 17.14% had postoperative complications (5 infections and 1 transient extremity paresis), and there were no deaths. Even in low-resource environments, effective and safe surgical care can be provided for drug-resistant epilepsy caused by temporal lobe disease. This study serves as a model of international collaboration and support for future hospitals in low-resource environments to replicate. [ABSTRACT FROM AUTHOR]

• Developed a novel two-step approach combining 1D convolutional neural networks (1DCNN) and long short-term memory (LSTM) networks to identify structural damages in bridges. • Utilized symbolic aggregate approximation (SAX) for efficient preprocessing and transformation of time-series data, enhancing data manageability and model training efficiency. • Implemented advanced data augmentation techniques to diversify time-series data, improving the model's robustness and generalization capabilities. • Demonstrated superior performance of the proposed 1DCNN-LSTM-Augmentation-SAX method over traditional models (1DCNN, LSTM, and 1DCNN-LSTM combinations) through validation on the chuong duong bridge structure. • Provided a comprehensive approach that leverages the strengths of CNNs in feature extraction and LSTMs in sequence learning for more accurate damage detection. This paper presents a novel two-step approach to identifying structural damages in bridge structure through the integration of 1D Convolutional Neural Network (1DCNN) and Long Short-Term Memory (LSTM) networks, enhanced by the augmentation and transformation techniques using Symbolic Aggregate approXimation (SAX) for time-series data analysis. In the first step, the time-series data of the bridge is diversified and quantified by augmentation techniques to make the model more robust and increase its generalization capabilities. After that, SAX is implemented to reduce the volume and categorize time series data through the transformation of continuous time series into discrete symbols, thereby decreasing the size of the data for more efficient training performance. In the second step, an advanced DL model combining 1DCNN and LSTM is proposed to tackle the damage identification problems of the processed data. By leveraging the strengths of CNNs in feature extraction and LSTMs in sequence learning, combined with advanced techniques for data augmentation, our methodology offers a robust solution not only for improving the model's training process but also for enabling it to learn from a more diverse and comprehensive dataset that mimics different damage scenarios, allowing more accurate detection of damages within bridge structures. Validation of the proposed method is conducted using time-series data collected from Chuong Duong Bridge structure. The effectiveness of the proposed method is compared with other models, such as 1DCNN, LSTM, and the combined 1DCNN-LSTM. The results show that the proposed 1DCNN-LSTM-SAX outperforms the other methods in terms of accuracy and, thus, can be used extensively to deal with the damage identification problems of bridges using time-series data. [ABSTRACT FROM AUTHOR]

Abstract: Text line segmentation in handwritten documents is an important step in document image processing. We present a new text line segmentation method based on the Mumford–Shah model. The algorithm is script independent. In addition, we use morphing to remove overlaps between neighboring text lines and connect broken ones. Experimental results show the validity of our method. [Copyright &y& Elsevier]

Abstract: For the first time, a genetic framework using contextual knowledge is proposed for segmentation and recognition of unconstrained handwritten numeral strings. New algorithms have been developed to locate feature points on the string image, and to generate possible segmentation hypotheses. A genetic representation scheme is utilized to show the space of all segmentation hypotheses (chromosomes). For the evaluation of segmentation hypotheses, a novel evaluation scheme is introduced, in order to improve the outlier resistance of the system. Our genetic algorithm tries to search and evolve the population of segmentation hypotheses, and to find the one with the highest segmentation/recognition confidence. The NIST NSTRING SD19 and CENPARMI databases were used to evaluate the performance of our proposed method. Our experiments showed that proper use of contextual knowledge in segmentation, evaluation and search greatly improves the overall performance of the system. On average, our system was able to obtain correct recognition rates of 95.28% and 96.42% on handwritten numeral strings using neural network and support vector classifiers, respectively. These results compare favorably with the ones reported in the literature. [Copyright &y& Elsevier]

Abstract: This paper studies the discrimination of similar handwritten numerals based on invariant curvature features. High-order B-splines are used to calculate the curvature of the contours of handwritten numerals. The concept of a distribution center is introduced so that a one-dimensional periodic signal can be normalized as shift invariant. Consequently, the curvature of the contour of a character becomes rotation invariant. To reduce the dimension of the features, wavelet basis decomposition is used to produce more compact features. Finally, artificial neural network (ANN) and support vector machines (SVM) are employed to train the features and design classifiers of high recognition rates. [Copyright &y& Elsevier]

• damage detection using changes in modal curvature and Convolutional Neural Network (CNN) • Gapped Smoothing Method (GSM) vibration-based damage detection method. • Application to Bo Nghi bridge. • The combination of GSM and CNN can be used for damage detection and localization. This paper addresses a damage detection method based on changes in modal curvature combined with Convolutional Neural Network (CNN). The use of modal curvature for damage detection is very well known. Some methods require modal data of healthy structures as a reference, others do not. Gapped Smoothing Method (GSM) is a vibration-based damage detection methodology, which makes use of modal curvature for identifying the location of structural damage and does not require information of the intact structure. The Bo Nghi bridge is used as an illustrative example. This bridge consists of four T-shaped concrete simply supported girders. One single beam with the same length and cross-section of the bridge girder is modeled and used to extract numerical data to train the CNN. A CNN is trained by using images from the damage index of the GSM to classify the damage location in the numerical beam model. Finally, the finite element model of the bridge is built and used to model damage scenarios to test the trained CNN. The results indicate that the combination of GSM and CNN can be used for damage detection and localization. [ABSTRACT FROM AUTHOR]

Highlights • The NTCP S267F variant is associated with a lower risk of chronic HBV infection. • First study on NTCP S267F variant on concomitant HDV infection in Vietnamese population. • The NTCP S267F is correlated with a reduced risk on clinical progression of liver cirrhosis and liver cancer. Abstract Objectives To determine potential associations of the rs2296651 variant (c.800C > T, S267F) of NTCP with HBV and HBV plus concomitant HDV infection as well as with the progression of related liver diseases. Methods The S267F variant was genotyped by DNA sequencing in 620 HBV-infected patients and 214 healthy controls (HCs). Among the patients, 450 individuals were tested for HDV by a nested PCR assay. Logistic regression was applied to examine the association. Results The S267F variant was found more frequently among HCs (16%) compared to HBV-infected (6%) and HBV-HDV co-infected patients (3%) (HBV patients vs HC: OR = 0.32, P = 0.00002 and HDV patients vs. HC: OR = 0.17, P = 0.018). The frequency of S267F variant was inversely correlated with CHB, LC or HCC patients compared with HCs (OR = 0.31, P = 0.001; OR = 0.32, P = 0.013; OR = 0.34, P = 0.002, respectively). S267F variant was also associated with decreased risk of the development of advanced liver cirrhosis (LC) and hepatocellular carcinoma (HCC) (Child B and C vs. Child A, OR = 0.26, adjusted P = 0.016; BCLC B,C,D vs. BCLC A, OR = 0.038, P = 0.045, respectively). In addition, patients with the genotype CT had lower levels of AST, ALT, total and direct bilirubin as well as higher platelet counts, indicating an association with a more favorable clinical outcome. Conclusion The NTCP S267F variant of the SLC10A1 gene exhibits protective effects against HBV and HDV infection and is associated with a reduced risk of developing to advanced stages of LC and HCC. [ABSTRACT FROM AUTHOR]

• FG-GPLRC circular plates and spherical caps are stiffened by spiderweb stiffeners. • Five distribution laws of GPLs of plate/cap and stiffeners are suggested. • Lekhnitskii's smeared stiffener technique is expanded for spiderweb stiffeners. • Donnell theory and the Ritz energy method are employed to obtain explicit results. • The postbuckling behavior of stiffened plates/caps is numerically investigated. For the first time, the problem of nonlinear postbuckling of circular plates and shallow spherical caps reinforced by meridian, parallel stiffeners, and spiderweb stiffeners based on the Donnell shell theory (DST) and von Kármán geometric nonlinearities is presented. The caps/plates and stiffeners are made from functionally graded graphene platelet-reinforced composite (FG-GPLRC). These stiffened structures are subjected to uniformly distributed external pressure or/and uniformly distributed thermal loads and are rested on a nonlinear elastic foundation. By expanding Lekhnitskii's smeared stiffener technique and employing the Ritz method of energy, the formulas to determine the postbuckling curves of the external pressure–deflection and thermal load-deflection relations of stiffened plates/spherical caps are derived. Meaningful discussions of the various influences of FG-GPLRC stiffeners, material distributions of plate/cap and stiffeners, and geometrical, material, and foundation parameters are shown in the content of the numerical investigations. [ABSTRACT FROM AUTHOR]

Highlights • A new connection named ECS connection at precast concrete beam end was developed. • Static bearing capacity experiments on twelve ECS connections were conducted. • The ultimate load of the ECS connection was investigated by parametric studies. • The optimum design of the ECS connection was proposed. Abstract Embedded channel-shaped steel (ECS) connection at precast concrete beam-end proposed in this paper is a kind of novel connection. To study the performance of ECS connection, experiments on 12 specimens were carried out. The embedded length of the ECS, the stirrup spacing, the cavity in the mid-span of the beam and the eccentricity of the ECS were taken into account. Two kinds of failure modes were observed: lever-out of ECS and crushing of the beam-end concrete. The experimental results indicate the following: (1) as the length of ECS increases, the bearing capacity of the ECS connection increases; (2) the stirrup densification could significantly improve the bearing and deformation capacity; (3) the mid-span cavity of the beam has limited influence on the bearing and deformation capacity of the connection; (4) the eccentricity of ECS could slightly reduce the bearing capacity owing to the additional torsion. Subsequently, finite element (FE) models were developed using ABAQUS and validated by the test results. The FE simulations show good agreement with the test results. Based on the verified FE models, a parametric study was conducted, considering different ECS properties including embedded length, section and eccentricity, various stirrup spacing at the beam-end, concrete grade and the distance between cavity and ECS. Finally, to prevent the lever-out of ECS, the optimum design of the ECS connection was proposed according to previous discussion. [ABSTRACT FROM AUTHOR]

• A coupled model between an artificial neural network (ANN) and antlion optimizer (ALO), so-called ALOANN is proposed. • Two numerical studies of the plate structures with single and multiple damage scenarios are presented. • Damage indicator coupled with ALOANN show better performance in localization and quantification damages. Mode shape-based method has shown its dominance in failure identification of beams. However, it is a challenge for fault detection in a plate structure. It requires combined information in two directions to determine the damage location. In this study, a damage index, namely mode shape derivative based damage identification (MSDBDI), is applied to localize damage in fixed-free plate structures. Two-dimensional (2D) displacement mode shapes and their derivatives are used to identify the MSDBDI index. It is a fact that this indicator is stuck in indicating the damage severity. Hence, a coupled model between an artificial neural network (ANN) and antlion optimizer (ALO), so-called ALOANN is used to overcome this drawback. In this method, ALO instead of a backpropagation algorithm is used to look for the best initial values of learnable parameters of ANN i.e. weights and biases through mean squared error (MSE). These obtained parameters are added to ANN for damage identification. The efficiency of the proposed approach is tested with two numerical studies of the plate structures with single and multiple damage scenarios. In the first application, damage scenarios in a plate-like structure are detected. In the second application, ALO first is used to build a FE model of a composite structure based on a vibration experiment. Then the slab of the updated model is assumed to be suffered several damage scenarios. In both applications, failures are localized by using damage index. Then, the proposed approach is used to quantify the corresponding extent by means of changes in frequencies and displacement mode shapes. Values of damage index are achieved from modal properties of one or three out of the first five modes of the two considered structures. A conventional ANN also is investigated for comparison. Results of damage identification indicate that the damage indicator coupled with ALOANN show better performance compared with using ANN alone even when a noise level is assigned to modal properties. [ABSTRACT FROM AUTHOR]

• Hybrid feedforward neural networks and marine predator algorithm for structural health monitoring. • Superiority of the algorithm is demonstrated through comparison with other algorithms such as PSO, GSA, PSOGSA, and GWO. • Application to a simply supported beam, a two-span continuous beam, and a laboratory free-free beam. Finite element (FE) based structural health monitoring (SHM) algorithms seek to update structural damage indices through solving an optimisation problem in which the difference between the response of the real structure and a corresponding FE model to some excitation force is minimised. These techniques, therefore, exploit advanced optimisation algorithms to alleviate errors stemming from the lack of information or the use of highly noisy measured responses. This study proposes an effective approach for damage detection by using a recently developed novel swarm intelligence algorithm, i.e. the marine predator algorithm (MPA). In the proposed approach, optimal foraging strategy and marine memory are employed to improve the learning ability of feedforward neural networks. After training, the hybrid feedforward neural networks and marine predator algorithm, MPAFNN, produces the best combination of connection weights and biases. These weights and biases then are re-input to the networks for prediction. Firstly, the classification capability of the proposed algorithm is investigated in comparison with some well-known optimization algorithms such as particle swarm optimization (PSO), gravitational search algorithm (GSA), hybrid particle swarm optimization-gravitational search algorithm (PSOGSA), and grey wolf optimizer (GWO) via four classification benchmark problems. The superior and stable performance of MPAFNN proves its effectiveness. Then, the proposed method is applied for damage identification of three numerical models, i.e. a simply supported beam, a two-span continuous beam, and a laboratory free-free beam by using modal flexibility indices. The obtained results reveal the feasibility of the proposed approach in damage identification not only for different structures with single damage and multiple damage, but also considering noise effect. [ABSTRACT FROM AUTHOR]

In this paper, we propose an efficient Artificial Neural Network (ANN) based on the global search capacity of evolutionary algorithms (EAs) to identify damages in laminated composite structures. With remarkable advances, ANN has taken off over the last decades. However, ANN also has major drawbacks relating to local minima issues because it applies backpropagation algorithms based on gradient descent (GD) techniques. This leads to a substantial reduction in the effectiveness and accuracy of ANN. Some researchers have been come up with some solutions to tackle the local minimal problems of ANN by looking for starting beneficial points to eliminate initial local minima based on the global search capacity of stochastic algorithms. Nevertheless, it is commonly acknowledged that those solutions are no longer useful or even counterproductive in some cases if the network contains too many local minima distributed deeply in the search space. Hence, we propose a novel approach applying the fast convergence speed of GD techniques of ANN and the global search capacity of EAs to train the network. The core idea is that EAs are employed to work parallel with ANN during the process of training the network. This guarantees that the network possibly determines the best solution fast and avoids getting stuck in local minima. To enhance the efficiency of the global search capacity, in this work, a hybrid metaheuristic optimization algorithm (HGACS) of EAs is also proposed, which possibly gains the advantages of both Genetic Algorithm (GA) and Cuckoo Search (CS). GA is applied to generate initial populations with the best quality derived from the ability of crossover and mutation operators, whereas CS with global search capacity is used to seek the best solution. Moreover, to deal with the large amount of data utilized to train the network, a vectorization technique is applied for the data of the objective function, which considerably decreases the computational cost. The obtained results prove that the proposed method is superior to traditional ANN, other hybrid-ANNs, and HGACS in terms of accuracy, and significantly reduces computational time compared with HGACS. [ABSTRACT FROM AUTHOR]

• Damage detection in structures by applying using artificial neural network (ANN) and cuckoo search (CS) algorithm. • Structural damage localization and quantification in bridges and beam-like structures. • Improve ANN training parameters using CS. • ANN combined with CS (ANN-CS) is accurate and cost effective. This paper presents a new approach for damage detection in structures by applying a flexible combination based on an artificial neural network (ANN) and cuckoo search (CS) algorithm. ANN has become one of the most powerful tools employing computational intelligence techniques to tackle complex problems in numerous fields. However, due to the application of backpropagation algorithms based on gradient descent, a major drawback of ANN is the common problem of local minima that acts as a great hindrance to the search for the best solution. To overcome this disadvantage, we propose to combine ANN with evolutionary algorithms based on global search techniques. This paper employs CS to improve ANN training parameters (weight and bias) by minimizing the difference between real and desired outputs and then using these parameters to generate the network. Two numerical models, comprising a steel beam calibrated using experimental measurements and a large-scale truss bridge, are used to assess the robustness of the proposed approach. The results demonstrate that ANN combined with CS (ANN-CS) is accurate and requires a lower computational time than ANN, and evolutionary algorithm (EA) alone in terms of structural damage localization and quantification. [ABSTRACT FROM AUTHOR]