scholarly journals USING APPLICATION OF AN ARTIFICIAL NEURAL NETWORK SYSTEM TO BACKCALCULATE PAVEMENT ELASTIC MODULUS

2020 ◽  
pp. 84-93
Author(s):  
M. M. M. Elshamy ◽  
A. N. Tiraturyan
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Elastic Moduli ◽  
Shock Loading ◽  
Back Propagation ◽  
Data Set ◽  
Neural Network System ◽  
Artificial Neural ◽  
Instrumental Recording ◽  
Accuracy Of Prediction

Statement of the problem. The article is devoted to the use of artificial neural networks in solving the problems of processing the results of instrumental recording of bowls of flexible pavement deflections using FWD shock loading settings. Results. The analysis was carried out, the shortcomings of the existing processing methods were noted, in particular the “backcalculation” method, which consists of a long calculation time, and the instability of the results obtained. The structure of the artificial neural network was built to determine the elastic moduli of the pavement layers. Training of an artificial neural network was carried out using the method of back propagation of error. Conclusions. The developed neural network has shown good results in training on the test data set, as well as high accuracy of prediction of the elastic moduli of the pavement.

Use of Artificial Neural Networks for Calculating Elasticity Modulus of Construction Layers of Non-Rigid Road Surfacing Based on Natural Data

2020 ◽  
pp. 111-120
Author(s):  
М. М. М. Елшами ◽  
А. Н. Тиратурян ◽  
А. Н. Канищев
Keyword(s):  
Neural Network ◽  
Neural Networks ◽  
Artificial Neural Network ◽  
Artificial Neural Networks ◽  
Elastic Moduli ◽  
Back Propagation ◽  
Data Set ◽  
Artificial Neural ◽  
Instrumental Recording ◽  
Accuracy Of Prediction

Постановка задачи. Рассматриваются вопросы использования искусственных нейронных сетей при решении задач обработки результатов инструментальных регистраций чаш прогибов нежесткой дорожной одежды с использованием установок ударного нагружения FWD . Результаты. Проведен анализ и отмечены недостатки существующих методов обработки экспериментальных чаш прогибов, в частности метода обратного расчета модулей упругости слоев дорожных одежд, заключающиеся в длительном времени выполнения расчетов и неустойчивости получаемых результатов. Построена структура искусственной нейронной сети для определения модулей упругости слоев дорожной одежды. Обучение искусственной нейронной сети осуществлялось с использованием метода обратного распространения ошибки. Выводы. Разработанная нейронная сеть продемонстрировала хорошие результаты при обучении по тестовому набору данных, а также высокую точность прогнозирования модулей упругости слоев дорожных одежд. Statement of the problem. The article is devoted to the use of artificial neural networks in solving the problems of processing the results of instrumental recording of bowls of deflections of non-rigid road surfacing using FWD shock loading settings. Results. The analysis was carried out, the shortcomings of the existing processing methods were identified, in particular the backcalculation method, which involves a long calculation time, and the instability of the results obtained. The structure of the artificial neural network was designed to determine the elastic moduli of the pavement layers. Training of an artificial neural network was carried out using the method of back propagation of error. Conclusions. The developed neural network has shown good results in training on the test data set, as well as high accuracy of prediction of the elastic moduli of the pavement.

scholarly journals Artificial Neural Network Modelling of Photodegradation in Suspension of Manganese Doped Zinc Oxide Nanoparticles under Visible-Light Irradiation

2014 ◽  
Vol 2014 ◽  
pp. 1-10 ◽  
Author(s):  
Yadollah Abdollahi ◽  
Azmi Zakaria ◽  
Nor Asrina Sairi ◽  
Khamirul Amin Matori ◽  
Hamid Reza Fard Masoumi ◽  
...  
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Visible Light ◽  
Visible Light Irradiation ◽  
Back Propagation ◽  
Light Irradiation ◽  
Maximum Efficiency ◽  
Data Set ◽  
Artificial Neural

The artificial neural network (ANN) modeling ofm-cresol photodegradation was carried out for determination of the optimum and importance values of the effective variables to achieve the maximum efficiency. The photodegradation was carried out in the suspension of synthesized manganese doped ZnO nanoparticles under visible-light irradiation. The input considered effective variables of the photodegradation were irradiation time, pH, photocatalyst amount, and concentration ofm-cresol while the efficiency was the only response as output. The performed experiments were designed into three data sets such as training, testing, and validation that were randomly splitted by the software’s option. To obtain the optimum topologies, ANN was trained by quick propagation (QP), Incremental Back Propagation (IBP), Batch Back Propagation (BBP), and Levenberg-Marquardt (LM) algorithms for testing data set. The topologies were determined by the indicator of minimized root mean squared error (RMSE) for each algorithm. According to the indicator, the QP-4-8-1, IBP-4-15-1, BBP-4-6-1, and LM-4-10-1 were selected as the optimized topologies. Among the topologies, QP-4-8-1 has presented the minimum RMSE and absolute average deviation as well as maximum R-squared. Therefore, QP-4-8-1 was selected as final model for validation test and navigation of the process. The model was used for determination of the optimum values of the effective variables by a few three-dimensional plots. The optimum points of the variables were confirmed by further validated experiments. Moreover, the model predicted the relative importance of the variables which showed none of them was neglectable in this work.

scholarly journals Artificial neural network classification-based skin cancer detection

2017 ◽  
Vol 7 (1.1) ◽  
pp. 591
Author(s):  
M. Shyamala Devi ◽  
A.N. Sruthi ◽  
P. Balamurugan
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Skin Cancer ◽  
Back Propagation ◽  
Modern Technology ◽  
Data Set ◽  
Processing Strategies ◽  
Skin Cancers ◽  
Artificial Neural ◽  
Artificial Neural Network Ann

At present, skin cancers are extremely the most severe and life-threatening kind of cancer. The majority of the pores and skin cancers are completely remediable at premature periods. Therefore, a premature recognition of pores and skin cancer can effectively protect the patients. Due to the progress of modern technology, premature recognition is very easy to identify. It is not extremely complicated to discover the affected pores and skin cancers with the exploitation of Artificial Neural Network (ANN). The treatment procedure exploits image processing strategies and Artificial Intelligence. It must be noted that, the dermoscopy photograph of pores and skin cancer is effectively determined and it is processed to several pre-processing for the purpose of noise eradication and enrichment in image quality. Subsequently, the photograph is distributed through image segmentation by means of thresholding. Few components distinctive for skin most cancers regions. These features are mined the practice of function extraction scheme - 2D Wavelet Transform scheme. These outcomes are provides to the Back-Propagation Neural (BPN) Network for effective classification. This completely categorizes the data set into either cancerous or non-cancerous. 

A Single Hidden Layer Artificial Neural Network Model that Employs Algebraic Neurons: Algebraic Learning Machine

2021 ◽  
Author(s):  
Ömer Faruk Ertuğrul
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Back Propagation ◽  
Mathematical Operation ◽  
Functional Link ◽  
Neural Network System ◽  
Benchmark Datasets ◽  
Artificial Neural ◽  
Learning Machine ◽  
Hidden Layer

Abstract Artificial neural networks (ANN) have been employed successfully because of their high modeling capability. Many versions of the ANN have been proposed to increase the modeling ability. Since ANN based on the biological neural network system, the only mathematical operation is summation or subtraction (while the coefficients are negative). This research was done to investigate the application of other mathematical operations, which are multiplication, division, logarithm, and exponential, in nodes. Based on this fact, a novel a single hidden layer feed-forward artificial neural network (SLFN) model, which was called was algebraic learning machine (ALM), was proposed. The proposed ALM was evaluated and validated with 60 different benchmark datasets. Obtained results were compared with results obtained by each of the extreme learning machine (ELM), randomized artificial neural network, and random vector functional link, and back-propagation trained SLFN methods. Achieved results show that the proposed method is successful enough to be employed in classification and regression.

scholarly journals Self-Improving Generative Artificial Neural Network for Pseudorehearsal Incremental Class Learning

Algorithms ◽  
2019 ◽  
Vol 12 (10) ◽  
pp. 206 ◽  
Author(s):  
Diego Mellado ◽  
Carolina Saavedra ◽  
Steren Chabert ◽  
Romina Torres ◽  
Rodrigo Salas
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Novelty Detection ◽  
Activation Function ◽  
Learning Sequence ◽  
Data Set ◽  
Detection Model ◽  
Neural Network System ◽  
Small Set ◽  
Artificial Neural

Deep learning models are part of the family of artificial neural networks and, as such, they suffer catastrophic interference when learning sequentially. In addition, the greater number of these models have a rigid architecture which prevents the incremental learning of new classes. To overcome these drawbacks, we propose the Self-Improving Generative Artificial Neural Network (SIGANN), an end-to-end deep neural network system which can ease the catastrophic forgetting problem when learning new classes. In this method, we introduce a novel detection model that automatically detects samples of new classes, and an adversarial autoencoder is used to produce samples of previous classes. This system consists of three main modules: a classifier module implemented using a Deep Convolutional Neural Network, a generator module based on an adversarial autoencoder, and a novelty-detection module implemented using an OpenMax activation function. Using the EMNIST data set, the model was trained incrementally, starting with a small set of classes. The results of the simulation show that SIGANN can retain previous knowledge while incorporating gradual forgetfulness of each learning sequence at a rate of about 7% per training step. Moreover, SIGANN can detect new classes that are hidden in the data with a median accuracy of 43 % and, therefore, proceed with incremental class learning.

Eye Diseases Classification Using Back Propagation Artificial Neural Network

2021 ◽  
Vol 39 (1B) ◽  
pp. 11-20
Author(s):  
Hanaa M. Ahmad ◽  
Shrooq R. Hameed
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Color Space ◽  
Back Propagation ◽  
Texture Features ◽  
Eye Diseases ◽  
Color Histogram ◽  
Data Set ◽  
Human Eye ◽  
Artificial Neural

A human eye is a vital organ responsible for a person's vision. So, the early detection of eye diseases is essential. The objective of this paper deals with diagnosing of seven different external eye diseases that can be recognized by a human eye. These diseases cause problems either in eye pupil, in sclera of eye or in both or in eyelid. Color histogram and texture features extraction techniques with classification technique are used to achieve the goal of diagnosing external eye diseases.  Hue Min Max Diff (HMMD) color space is used to extract color histogram and texture features which were fed to Back Propagation Artificial Neural Network (BPANN) for classification. The comparative study states that the features extracted from HMMD color space is better than other features like Histogram of Oriented Gradient (HOG) features and give the same accuracy as features extracted directly from medical expert recorded symptoms. The proposed method is applied on external eye diseases data set consisting of 416 images with an accuracy rate of 85.26315%, which is the major result that was achieved in this study.

Performance Prediction of Mechanical Draft Wet Cooling Tower Using Artificial Neural Network

2014 ◽  
Vol 1070-1072 ◽  
pp. 1994-1997
Author(s):  
Zhe Tian Xu ◽  
Jia Chen Mao ◽  
Yi Qun Pan ◽  
Zhi Zhong Huang
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Cooling Tower ◽  
Back Propagation ◽  
Back Propagation Neural Network ◽  
Data Set ◽  
Artificial Neural ◽  
Artificial Neural Network Ann ◽  
Experimental Values ◽  
Prediction Approach

This paper proposed a prediction approach for the performance of the mechanical draft wet cooling tower based on artificial neural network (ANN). The inlet water temperature, the ambient wet bulb temperature and the ratio of water to air mass flow rate in the cooling tower were selected as the input parameters of a four-layer back propagation neural network (BPNN) to predict the temperature of the water at the tower outlet. After the test of the available data set, the BPNN results in a correlation coefficient of 0.9 between the predicted and experimental values. Thus the prediction performance is good and such prediction approach proves to be feasible and effective.

Austenite Formation Temperature Prediction in Steels Using an Artificial Neural Network

2008 ◽  
Vol 273-276 ◽  
pp. 335-341 ◽  
Author(s):  
M. Arjomandi ◽  
S.H. Sadati ◽  
H. Khorsand ◽  
H. Abdoos
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Chemical Composition ◽  
Empirical Equation ◽  
Learning Algorithm ◽  
Back Propagation ◽  
Austenite Formation ◽  
Data Set ◽  
Artificial Neural ◽  
Experimental Values

Determination of the temperature at which Austenite is formed is one of the important parameters in the heat treatment process. Chemical composition is an effective factor on these temperatures, particularly in steels that are used in various industries. In this research we have made an attempt to determine these temperatures based on the chemical composition of the steel. The technique used for this purpose is feedforward Artificial Neural Network (ANN) with the Back Propagation (BP) learning algorithm. A comparison is made between Ac1, Ac3 temperatures predicted with this model and those from the empirical equation as well as the experimental values obtained from costly and time-consuming tests in scientific and industrial centers for various steels. This comparison indicates that at Ac1, a better agreement exists between the ANN-predicted results and experimental values than the results from the empirical equation and experimental values. At Ac3, the results from the empirical equation are closer to those of the experimental than those predicted from the ANN. This was due to the dispersion of the data set used.

scholarly journals Self-Improving Generative Artificial Neural Network for Pseudo-Rehearsal Incremental Class Learning

2019 ◽  
Author(s):  
Diego Mellado ◽  
Carolina Saavedra ◽  
Steren Chabert ◽  
Romina Torres ◽  
Rodrigo Salas
Keyword(s):  
Neural Network ◽  
Artificial Neural Network ◽  
Novelty Detection ◽  
Activation Function ◽  
Learning Sequence ◽  
Data Set ◽  
Detection Model ◽  
Neural Network System ◽  
Small Set ◽  
Artificial Neural

Deep learning models are part of the family of artificial neural networks and, as such, it suffers of catastrophic interference when they learn sequentially. In addition, most of these models have a rigid architecture which prevents the incremental learning of new classes. To overcome these drawbacks, in this article we propose the Self-Improving Generative Artificial Neural Network (SIGANN), a type of end-to-end Deep Neural Network system which is able to ease the catastrophic forgetting problem when leaning new classes. In this method, we introduce a novelty detection model to automatically detect samples of new classes, moreover an adversarial auto-encoder is used to produce samples of previous classes. This system consists of three main modules: a classifier module implemented using a Deep Convolutional Neural Network, a generator module based on an adversarial autoencoder; and a novelty detection module, implemented using an OpenMax activation function. Using the EMNIST data set, the model was trained incrementally, starting with a small set of classes. The results of the simulation show that SIGANN is able to retain previous knowledge with a gradual forgetfulness for each learning sequence. Moreover, SIGANN can detect new classes that are hidden in the data and, therefore, proceed with incremental class learning.

Random Forest (RF) and Artificial Neural Network (ANN) Algorithms for LULC Mapping

2020 ◽  
Vol 38 (4A) ◽  
pp. 510-514
Author(s):  
Tay H. Shihab ◽  
Amjed N. Al-Hameedawi ◽  
Ammar M. Hamza
Keyword(s):  
Neural Network ◽  
Remote Sensing ◽  
Artificial Neural Network ◽  
Random Forest ◽  
Satellite Image ◽  
Landsat 8 ◽  
Optical Remote Sensing ◽  
Data Set ◽  
Artificial Neural ◽  
Artificial Neural Network Ann

In this paper to make use of complementary potential in the mapping of LULC spatial data is acquired from LandSat 8 OLI sensor images are taken in 2019.  They have been rectified, enhanced and then classified according to Random forest (RF) and artificial neural network (ANN) methods. Optical remote sensing images have been used to get information on the status of LULC classification, and extraction details. The classification of both satellite image types is used to extract features and to analyse LULC of the study area. The results of the classification showed that the artificial neural network method outperforms the random forest method. The required image processing has been made for Optical Remote Sensing Data to be used in LULC mapping, include the geometric correction, Image Enhancements, The overall accuracy when using the ANN methods 0.91 and the kappa accuracy was found 0.89 for the training data set. While the overall accuracy and the kappa accuracy of the test dataset were found 0.89 and 0.87 respectively.

Sign in / Sign up

Export Citation Format

Share Document