Displaying publications 1 - 20 of 409 in total

Abstract:
Sort:
  1. Wang X, Song T, Gong F, Zheng P
    Sci Rep, 2016 06 10;6:27624.
    PMID: 27283843 DOI: 10.1038/srep27624
    Neural-like computing models are versatile computing mechanisms in the field of artificial intelligence. Spiking neural P systems (SN P systems for short) are one of the recently developed spiking neural network models inspired by the way neurons communicate. The communications among neurons are essentially achieved by spikes, i. e. short electrical pulses. In terms of motivation, SN P systems fall into the third generation of neural network models. In this study, a novel variant of SN P systems, namely SN P systems with self-organization, is introduced, and the computational power of the system is investigated and evaluated. It is proved that SN P systems with self-organization are capable of computing and accept the family of sets of Turing computable natural numbers. Moreover, with 87 neurons the system can compute any Turing computable recursive function, thus achieves Turing universality. These results demonstrate promising initiatives to solve an open problem arisen by Gh Păun.
    Matched MeSH terms: Neural Networks (Computer)*
  2. Illias HA, Chai XR, Abu Bakar AH, Mokhlis H
    PLoS One, 2015;10(6):e0129363.
    PMID: 26103634 DOI: 10.1371/journal.pone.0129363
    It is important to predict the incipient fault in transformer oil accurately so that the maintenance of transformer oil can be performed correctly, reducing the cost of maintenance and minimise the error. Dissolved gas analysis (DGA) has been widely used to predict the incipient fault in power transformers. However, sometimes the existing DGA methods yield inaccurate prediction of the incipient fault in transformer oil because each method is only suitable for certain conditions. Many previous works have reported on the use of intelligence methods to predict the transformer faults. However, it is believed that the accuracy of the previously proposed methods can still be improved. Since artificial neural network (ANN) and particle swarm optimisation (PSO) techniques have never been used in the previously reported work, this work proposes a combination of ANN and various PSO techniques to predict the transformer incipient fault. The advantages of PSO are simplicity and easy implementation. The effectiveness of various PSO techniques in combination with ANN is validated by comparison with the results from the actual fault diagnosis, an existing diagnosis method and ANN alone. Comparison of the results from the proposed methods with the previously reported work was also performed to show the improvement of the proposed methods. It was found that the proposed ANN-Evolutionary PSO method yields the highest percentage of correct identification for transformer fault type than the existing diagnosis method and previously reported works.
    Matched MeSH terms: Neural Networks (Computer)*
  3. Marto A, Hajihassani M, Armaghani DJ, Mohamad ET, Makhtar AM
    ScientificWorldJournal, 2014;2014:643715.
    PMID: 25147856 DOI: 10.1155/2014/643715
    Flyrock is one of the major disturbances induced by blasting which may cause severe damage to nearby structures. This phenomenon has to be precisely predicted and subsequently controlled through the changing in the blast design to minimize potential risk of blasting. The scope of this study is to predict flyrock induced by blasting through a novel approach based on the combination of imperialist competitive algorithm (ICA) and artificial neural network (ANN). For this purpose, the parameters of 113 blasting operations were accurately recorded and flyrock distances were measured for each operation. By applying the sensitivity analysis, maximum charge per delay and powder factor were determined as the most influential parameters on flyrock. In the light of this analysis, two new empirical predictors were developed to predict flyrock distance. For a comparison purpose, a predeveloped backpropagation (BP) ANN was developed and the results were compared with those of the proposed ICA-ANN model and empirical predictors. The results clearly showed the superiority of the proposed ICA-ANN model in comparison with the proposed BP-ANN model and empirical approaches.
    Matched MeSH terms: Neural Networks (Computer)*
  4. Khan NA, Ibrahim Khalaf O, Andrés Tavera Romero C, Sulaiman M, Bakar MA
    Comput Intell Neurosci, 2022;2022:2710576.
    PMID: 35096038 DOI: 10.1155/2022/2710576
    In this study, the intelligent computational strength of neural networks (NNs) based on the backpropagated Levenberg-Marquardt (BLM) algorithm is utilized to investigate the numerical solution of nonlinear multiorder fractional differential equations (FDEs). The reference data set for the design of the BLM-NN algorithm for different examples of FDEs are generated by using the exact solutions. To obtain the numerical solutions, multiple operations based on training, validation, and testing on the reference data set are carried out by the design scheme for various orders of FDEs. The approximate solutions by the BLM-NN algorithm are compared with analytical solutions and performance based on mean square error (MSE), error histogram (EH), regression, and curve fitting. This further validates the accuracy, robustness, and efficiency of the proposed algorithm.
    Matched MeSH terms: Neural Networks (Computer)*
  5. Tunc C
    Sains Malaysiana, 2011;40:1455-1459.
    The main purpose of this paper is to introduce some new instability theorems related to certain fourth and fifth order nonlinear differential equations with a constant delay. By means of the Lyapunov-Krasovskii functional approach, we obtained two new results on the topic.
  6. Gong C, Xue B, Jing C, He CH, Wu GC, Lei B, et al.
    Math Biosci Eng, 2022 Sep 13;19(12):13276-13293.
    PMID: 36654046 DOI: 10.3934/mbe.2022621
    Brain community detection is an efficient method to represent the communities of brain networks. However, time-variable functions of the brain and the intricate brain community structure impose a great challenge on it. In this paper, a time-sequential graph adversarial learning (TGAL) framework is proposed to detect brain communities and characterize the structure of communities from brain networks. In the framework, a novel time-sequential graph neural network is designed as an encoder to extract efficient graph representations by spatio-temporal attention mechanism. Since it is difficult to capture the community structure, the measurable modularity loss is used to optimize by maximizing the modularity of the community. In addition, the framework employs an adversarial scheme to guide the learning of representation. The effectiveness of our model is shown through experiments on the real-world brain network datasets, and the great performance of brain community detection demonstrates the advantage of the proposed framework.
  7. Sohail MT, Yang M, Maresova P, Mustafa S
    Front Public Health, 2022;10:1046780.
    PMID: 36339186 DOI: 10.3389/fpubh.2022.1046780
    This study was conducted to evaluate public awareness about COVID with aimed to check public strategies against COVID-19. A semi structured questionnaire was collected and the data was analyzed using some statistical tools (PLS-SEM) and artificial neural networks (ANN). We started by looking at the known causal linkages between the different variables to see if they matched up with the hypotheses that had been proposed. Next, for this reason, we ran a 5,000-sample bootstrapping test to assess how strongly our findings corroborated the null hypothesis. PLS-SEM direct path analysis revealed HRP -> PA-COVID, HI -> PA-COVID, MU -> PA-COVID, PM -> PA-COVID, SD -> PA-COVID. These findings provide credence to the acceptance of hypotheses H1, H3, and H5, but reject hypothesis H2. We have also examined control factors such as respondents' age, gender, and level of education. Age was found to have a positive correlation with PA-COVID, while mean gender and education level were found to not correlate at all with PA-COVID. However, age can be a useful control variable, as a more seasoned individual is likely to have a better understanding of COVID and its effects on independent variables. Study results revealed a small moderation effect in the relationships between understudy independent and dependent variables. Education significantly moderates the relationship of PA-COVID associated with MU, PH, SD, RP, PM, PA-COVID, depicts the moderation role of education on the relationship between MU*Education->PA-COVID, HI*Education->PA.COVID, SD*Education->PA.COVID, HRP*Education->PA.COVID, PM*Education -> PA.COVID. The artificial neural network (ANN) model we've developed for spreading information about COVID-19 (PA-COVID) follows in the footsteps of previous studies. The root means the square of the errors (RMSE). Validity measures how well a model can predict a certain result. With RMSE values of 0.424 for training and 0.394 for testing, we observed that our ANN model for public awareness of COVID-19 (PA-COVID) had a strong predictive ability. Based on the sensitivity analysis results, we determined that PA. COVID had the highest relative normalized relevance for our sample (100%). These factors were then followed by MU (54.6%), HI (11.1%), SD (100.0%), HRP (28.5%), and PM (64.6%) were likewise shown to be the least important factors for consumers in developing countries struggling with diseases caused by contaminated water. In addition, a specific approach was used to construct a goodness-of-fit coefficient to evaluate the performance of the ANN models. The study will aid in the implementation of effective monitoring and public policies to promote the health of local people.
  8. Ali SS, Moinuddin M, Raza K, Adil SH
    ScientificWorldJournal, 2014;2014:850189.
    PMID: 24987745 DOI: 10.1155/2014/850189
    Radial basis function neural networks are used in a variety of applications such as pattern recognition, nonlinear identification, control and time series prediction. In this paper, the learning algorithm of radial basis function neural networks is analyzed in a feedback structure. The robustness of the learning algorithm is discussed in the presence of uncertainties that might be due to noisy perturbations at the input or to modeling mismatch. An intelligent adaptation rule is developed for the learning rate of RBFNN which gives faster convergence via an estimate of error energy while giving guarantee to the l 2 stability governed by the upper bounding via small gain theorem. Simulation results are presented to support our theoretical development.
    Matched MeSH terms: Neural Networks (Computer)*
  9. Ahmed AU, Islam MT, Ismail M, Kibria S, Arshad H
    ScientificWorldJournal, 2014;2014:253787.
    PMID: 25133214 DOI: 10.1155/2014/253787
    An artificial neural network (ANN) and affinity propagation (AP) algorithm based user categorization technique is presented. The proposed algorithm is designed for closed access femtocell network. ANN is used for user classification process and AP algorithm is used to optimize the ANN training process. AP selects the best possible training samples for faster ANN training cycle. The users are distinguished by using the difference of received signal strength in a multielement femtocell device. A previously developed directive microstrip antenna is used to configure the femtocell device. Simulation results show that, for a particular house pattern, the categorization technique without AP algorithm takes 5 indoor users and 10 outdoor users to attain an error-free operation. While integrating AP algorithm with ANN, the system takes 60% less training samples reducing the training time up to 50%. This procedure makes the femtocell more effective for closed access operation.
    Matched MeSH terms: Neural Networks (Computer)*
  10. Mohseni SA, Tan AH
    IEEE Trans Syst Man Cybern B Cybern, 2012 Dec;42(6):1645-53.
    PMID: 22665508 DOI: 10.1109/TSMCB.2012.2197610
    This paper proposes a new mixed training algorithm consisting of error backpropagation (EBP) and variable structure systems (VSSs) to optimize parameter updating of neural networks. For the optimization of the number of neurons in the hidden layer, a new term based on the output of the hidden layer is added to the cost function as a penalty term to make optimal use of hidden units related to weights corresponding to each unit in the hidden layer. VSS is used to control the dynamic model of the training process, whereas EBP attempts to minimize the cost function. In addition to the analysis of the imposed dynamics of the EBP technique, the global stability of the mixed training methodology and constraints on the design parameters are considered. The advantages of the proposed technique are guaranteed convergence, improved robustness, and lower sensitivity to initial weights of the neural network.
    Matched MeSH terms: Neural Networks (Computer)*
  11. Chan Phooi M'ng J, Mehralizadeh M
    PLoS One, 2016;11(6):e0156338.
    PMID: 27248692 DOI: 10.1371/journal.pone.0156338
    The motivation behind this research is to innovatively combine new methods like wavelet, principal component analysis (PCA), and artificial neural network (ANN) approaches to analyze trade in today's increasingly difficult and volatile financial futures markets. The main focus of this study is to facilitate forecasting by using an enhanced denoising process on market data, taken as a multivariate signal, in order to deduct the same noise from the open-high-low-close signal of a market. This research offers evidence on the predictive ability and the profitability of abnormal returns of a new hybrid forecasting model using Wavelet-PCA denoising and ANN (named WPCA-NN) on futures contracts of Hong Kong's Hang Seng futures, Japan's NIKKEI 225 futures, Singapore's MSCI futures, South Korea's KOSPI 200 futures, and Taiwan's TAIEX futures from 2005 to 2014. Using a host of technical analysis indicators consisting of RSI, MACD, MACD Signal, Stochastic Fast %K, Stochastic Slow %K, Stochastic %D, and Ultimate Oscillator, empirical results show that the annual mean returns of WPCA-NN are more than the threshold buy-and-hold for the validation, test, and evaluation periods; this is inconsistent with the traditional random walk hypothesis, which insists that mechanical rules cannot outperform the threshold buy-and-hold. The findings, however, are consistent with literature that advocates technical analysis.
    Matched MeSH terms: Neural Networks (Computer)*
  12. Abdelhaq M, Alsaqour R, Abdelhaq S
    PLoS One, 2015;10(5):e0120715.
    PMID: 25946001 DOI: 10.1371/journal.pone.0120715
    A mobile ad hoc network (MANET) is a set of mobile, decentralized, and self-organizing nodes that are used in special cases, such as in the military. MANET properties render the environment of this network vulnerable to different types of attacks, including black hole, wormhole and flooding-based attacks. Flooding-based attacks are one of the most dangerous attacks that aim to consume all network resources and thus paralyze the functionality of the whole network. Therefore, the objective of this paper is to investigate the capability of a danger theory-based artificial immune algorithm called the mobile dendritic cell algorithm (MDCA) to detect flooding-based attacks in MANETs. The MDCA applies the dendritic cell algorithm (DCA) to secure the MANET with additional improvements. The MDCA is tested and validated using Qualnet v7.1 simulation tool. This work also introduces a new simulation module for a flooding attack called the resource consumption attack (RCA) using Qualnet v7.1. The results highlight the high efficiency of the MDCA in detecting RCAs in MANETs.
    Matched MeSH terms: Neural Networks (Computer)*
  13. Siavash NK, Ghobadian B, Najafi G, Rohani A, Tavakoli T, Mahmoodi E, et al.
    Environ Res, 2021 05;196:110434.
    PMID: 33166537 DOI: 10.1016/j.envres.2020.110434
    Wind power is one of the most popular sources of renewable energies with an ideal extractable value that is limited to 0.593 known as the Betz-Joukowsky limit. As the generated power of wind machines is proportional to cubic wind speed, therefore it is logical that a small increment in wind speed will result in significant growth in generated power. Shrouding a wind turbine is an ordinary way to exceed the Betz limit, which accelerates the wind flow through the rotor plane. Several layouts of shrouds are developed by researchers. Recently an innovative controllable duct is developed by the authors of this work that can vary the shrouding angle, so its performance is different in each opening angle. As a wind tunnel investigation is heavily time-consuming and has a high cost, therefore just four different opening angles have been assessed. In this work, the performance of the turbine was predicted using multiple linear regression and an artificial neural network in a wide range of duct opening angles. For the turbine power generation and its rotor angular speed in different wind velocities and duct opening angles, regression and an ANN are suggested. The developed neural network model is found to possess better performance than the regression model for both turbine power curve and rotor speed estimation. This work revealed that in higher ranges of wind velocity, the turbine performance intensively will be a function of shrouding angle. This model can be used as a lookup table in controlling the turbines equipped with the proposed mechanism.
    Matched MeSH terms: Neural Networks (Computer)*
  14. Abdollahi A, Pradhan B
    Sensors (Basel), 2021 Jul 11;21(14).
    PMID: 34300478 DOI: 10.3390/s21144738
    Urban vegetation mapping is critical in many applications, i.e., preserving biodiversity, maintaining ecological balance, and minimizing the urban heat island effect. It is still challenging to extract accurate vegetation covers from aerial imagery using traditional classification approaches, because urban vegetation categories have complex spatial structures and similar spectral properties. Deep neural networks (DNNs) have shown a significant improvement in remote sensing image classification outcomes during the last few years. These methods are promising in this domain, yet unreliable for various reasons, such as the use of irrelevant descriptor features in the building of the models and lack of quality in the labeled image. Explainable AI (XAI) can help us gain insight into these limits and, as a result, adjust the training dataset and model as needed. Thus, in this work, we explain how an explanation model called Shapley additive explanations (SHAP) can be utilized for interpreting the output of the DNN model that is designed for classifying vegetation covers. We want to not only produce high-quality vegetation maps, but also rank the input parameters and select appropriate features for classification. Therefore, we test our method on vegetation mapping from aerial imagery based on spectral and textural features. Texture features can help overcome the limitations of poor spectral resolution in aerial imagery for vegetation mapping. The model was capable of obtaining an overall accuracy (OA) of 94.44% for vegetation cover mapping. The conclusions derived from SHAP plots demonstrate the high contribution of features, such as Hue, Brightness, GLCM_Dissimilarity, GLCM_Homogeneity, and GLCM_Mean to the output of the proposed model for vegetation mapping. Therefore, the study indicates that existing vegetation mapping strategies based only on spectral characteristics are insufficient to appropriately classify vegetation covers.
    Matched MeSH terms: Neural Networks (Computer)*
  15. Koay HV, Chuah JH, Chow CO, Chang YL, Rudrusamy B
    Sensors (Basel), 2021 Jul 15;21(14).
    PMID: 34300577 DOI: 10.3390/s21144837
    Distracted driving is the prime factor of motor vehicle accidents. Current studies on distraction detection focus on improving distraction detection performance through various techniques, including convolutional neural networks (CNNs) and recurrent neural networks (RNNs). However, the research on detection of distracted drivers through pose estimation is scarce. This work introduces an ensemble of ResNets, which is named Optimally-weighted Image-Pose Approach (OWIPA), to classify the distraction through original and pose estimation images. The pose estimation images are generated from HRNet and ResNet. We use ResNet101 and ResNet50 to classify the original images and the pose estimation images, respectively. An optimum weight is determined through grid search method, and the predictions from both models are weighted through this parameter. The experimental results show that our proposed approach achieves 94.28% accuracy on AUC Distracted Driver Dataset.
    Matched MeSH terms: Neural Networks (Computer)*
  16. Hai A, Bharath G, Daud M, Rambabu K, Ali I, Hasan SW, et al.
    Chemosphere, 2021 Nov;283:131162.
    PMID: 34157626 DOI: 10.1016/j.chemosphere.2021.131162
    Pyrolysis of agricultural biomass is a promising technique for producing renewable energy and effectively managing solid waste. In this study, groundnut shell (GNS) was processed at 500 °C in an inert gas atmosphere with a gas flow rate and a heating rate of 10 mL/min and 10 °C/min, respectively, in a custom-designed fluidized bed pyrolytic-reactor. Under optimal operating conditions, the GNS-derived pyrolytic-oil yield was 62.8 wt.%, with the corresponding biochar (19.5 wt.%) and biogas yields (17.7 wt.%). The GC-MS analysis of the GNS-based bio-oil confirmed the presence of (trifluoromethyl)pyridin-2-amine (18.814%), 2-Fluoroformyl-3,3,4,4-tetrafluoro-1,2-oxazetidine (16.23%), 5,7-dimethyl-1H-Indazole (11.613%), N-methyl-N-nitropropan-2-amine (6.5%) and butyl piperidino sulfone (5.668%) as major components, which are used as building blocks in the biofuel, pharmaceutical, and food industries. Furthermore, a 2 × 5 × 1 artificial neural network (ANN) architecture was developed to predict the decomposition behavior of GNS at heating rates of 5, 10, and 20 °C/min, while the thermodynamic and kinetic parameters were estimated using a non-isothermal model-free method. The Popescu method predicted activation energy (Ea) of GNS biomass ranging from 111 kJ/mol to 260 kJ/mol, with changes in enthalpy (ΔH), Gibbs-free energy (ΔG), and entropy (ΔS) ranging from 106 to 254 kJ/mol, 162-241 kJ/mol, and -0.0937 to 0.0598 kJ/mol/K, respectively. The extraction of high-quality precursors from GNS pyrolysis was demonstrated in this study, as well as the usefulness of the ANN technique for thermogravimetric analysis of biomass.
    Matched MeSH terms: Neural Networks (Computer)*
  17. Saeed AQ, Sheikh Abdullah SNH, Che-Hamzah J, Abdul Ghani AT
    J Med Internet Res, 2021 09 21;23(9):e27414.
    PMID: 34236992 DOI: 10.2196/27414
    BACKGROUND: Glaucoma leads to irreversible blindness. Globally, it is the second most common retinal disease that leads to blindness, slightly less common than cataracts. Therefore, there is a great need to avoid the silent growth of this disease using recently developed generative adversarial networks (GANs).

    OBJECTIVE: This paper aims to introduce a GAN technology for the diagnosis of eye disorders, particularly glaucoma. This paper illustrates deep adversarial learning as a potential diagnostic tool and the challenges involved in its implementation. This study describes and analyzes many of the pitfalls and problems that researchers will need to overcome to implement this kind of technology.

    METHODS: To organize this review comprehensively, articles and reviews were collected using the following keywords: ("Glaucoma," "optic disc," "blood vessels") and ("receptive field," "loss function," "GAN," "Generative Adversarial Network," "Deep learning," "CNN," "convolutional neural network" OR encoder). The records were identified from 5 highly reputed databases: IEEE Xplore, Web of Science, Scopus, ScienceDirect, and PubMed. These libraries broadly cover the technical and medical literature. Publications within the last 5 years, specifically 2015-2020, were included because the target GAN technique was invented only in 2014 and the publishing date of the collected papers was not earlier than 2016. Duplicate records were removed, and irrelevant titles and abstracts were excluded. In addition, we excluded papers that used optical coherence tomography and visual field images, except for those with 2D images. A large-scale systematic analysis was performed, and then a summarized taxonomy was generated. Furthermore, the results of the collected articles were summarized and a visual representation of the results was presented on a T-shaped matrix diagram. This study was conducted between March 2020 and November 2020.

    RESULTS: We found 59 articles after conducting a comprehensive survey of the literature. Among the 59 articles, 30 present actual attempts to synthesize images and provide accurate segmentation/classification using single/multiple landmarks or share certain experiences. The other 29 articles discuss the recent advances in GANs, do practical experiments, and contain analytical studies of retinal disease.

    CONCLUSIONS: Recent deep learning techniques, namely GANs, have shown encouraging performance in retinal disease detection. Although this methodology involves an extensive computing budget and optimization process, it saturates the greedy nature of deep learning techniques by synthesizing images and solves major medical issues. This paper contributes to this research field by offering a thorough analysis of existing works, highlighting current limitations, and suggesting alternatives to support other researchers and participants in further improving and strengthening future work. Finally, new directions for this research have been identified.

    Matched MeSH terms: Neural Networks (Computer)*
  18. Nisar K, Sabir Z, Zahoor Raja MA, Ibrahim AAA, Mahmoud SR, Balubaid M, et al.
    Sensors (Basel), 2021 Sep 30;21(19).
    PMID: 34640887 DOI: 10.3390/s21196567
    In this study, the numerical computation heuristic of the environmental and economic system using the artificial neural networks (ANNs) structure together with the capabilities of the heuristic global search genetic algorithm (GA) and the quick local search interior-point algorithm (IPA), i.e., ANN-GA-IPA. The environmental and economic system is dependent of three categories, execution cost of control standards and new technical diagnostics elimination costs of emergencies values and the competence of the system of industrial elements. These three elements form a nonlinear differential environmental and economic system. The optimization of an error-based objective function is performed using the differential environmental and economic system and its initial conditions. The optimization of an error-based objective function is performed using the differential environmental and economic system and its initial conditions.
    Matched MeSH terms: Neural Networks (Computer)*
  19. Hentabli H, Bengherbia B, Saeed F, Salim N, Nafea I, Toubal A, et al.
    Int J Mol Sci, 2022 Oct 30;23(21).
    PMID: 36362018 DOI: 10.3390/ijms232113230
    Determining and modeling the possible behaviour and actions of molecules requires investigating the basic structural features and physicochemical properties that determine their behaviour during chemical, physical, biological, and environmental processes. Computational approaches such as machine learning methods are alternatives to predicting the physiochemical properties of molecules based on their structures. However, the limited accuracy and high error rates of such predictions restrict their use. In this paper, a novel technique based on a deep learning convolutional neural network (CNN) for the prediction of chemical compounds' bioactivity is proposed and developed. The molecules are represented in the new matrix format Mol2mat, a molecular matrix representation adapted from the well-known 2D-fingerprint descriptors. To evaluate the performance of the proposed methods, a series of experiments were conducted using two standard datasets, namely the MDL Drug Data Report (MDDR) and Sutherland, datasets comprising 10 homogeneous and 14 heterogeneous activity classes. After analysing the eight fingerprints, all the probable combinations were investigated using the five best descriptors. The results showed that a combination of three fingerprints, ECFP4, EPFP4, and ECFC4, along with a CNN activity prediction process, achieved the highest performance of 98% AUC when compared to the state-of-the-art ML algorithms NaiveB, LSVM, and RBFN.
    Matched MeSH terms: Neural Networks (Computer)*
  20. Al-Saffar A, Awang S, Al-Saiagh W, Al-Khaleefa AS, Abed SA
    Sensors (Basel), 2021 Nov 02;21(21).
    PMID: 34770612 DOI: 10.3390/s21217306
    Handwriting recognition refers to recognizing a handwritten input that includes character(s) or digit(s) based on an image. Because most applications of handwriting recognition in real life contain sequential text in various languages, there is a need to develop a dynamic handwriting recognition system. Inspired by the neuroevolutionary technique, this paper proposes a Dynamically Configurable Convolutional Recurrent Neural Network (DC-CRNN) for the handwriting recognition sequence modeling task. The proposed DC-CRNN is based on the Salp Swarm Optimization Algorithm (SSA), which generates the optimal structure and hyperparameters for Convolutional Recurrent Neural Networks (CRNNs). In addition, we investigate two types of encoding techniques used to translate the output of optimization to a CRNN recognizer. Finally, we proposed a novel hybridized SSA with Late Acceptance Hill-Climbing (LAHC) to improve the exploitation process. We conducted our experiments on two well-known datasets, IAM and IFN/ENIT, which include both the Arabic and English languages. The experimental results have shown that LAHC significantly improves the SSA search process. Therefore, the proposed DC-CRNN outperforms the handcrafted CRNN methods.
    Matched MeSH terms: Neural Networks (Computer)*
Filters
Contact Us

Please provide feedback to Administrator (afdal@afpm.org.my)

External Links