Effectiveness Of Artificial Neural Networks Research Articles

Effectiveness of Artificial Neural Networks for Solving Inverse Problems in Magnetic Field-Based Localization.

Recently, indoor localization has become an active area of research. Although there are various approaches to indoor localization, methods that utilize artificially generated magnetic fields from a target device are considered to be the best in terms of localization accuracy under non-line-of-sight conditions. In magnetic field-based localization, the target position must be calculated based on the magnetic field information detected by multiple sensors. The calculation process is equivalent to solving a nonlinear inverse problem. Recently, a machine-learning approach has been proposed to solve the inverse problem. Reportedly, adopting the k-nearest neighbor algorithm (k-NN) enabled the machine-learning approach to achieve fairly good performance in terms of both localization accuracy and computational speed. Moreover, it has been suggested that the localization accuracy can be further improved by adopting artificial neural networks (ANNs) instead of k-NN. However, the effectiveness of ANNs has not yet been demonstrated. In this study, we thoroughly investigated the effectiveness of ANNs for solving the inverse problem of magnetic field-based localization in comparison with k-NN. We demonstrate that despite taking longer to train, ANNs are superior to k-NN in terms of localization accuracy. The k-NN is still valid for predicting fairly accurate target positions within limited training times.

Evaluation of the Effectiveness of Artificial Neural Network Based on Correcting Scoliosis and Improving Spinal Health in University Students.

In recent years, spinal disorders, spinal deformities, and scoliosis have become more frequent, especially chronic systemic inflammatory diseases, such as early-onset scoliosis, idiopathic scoliosis, and ankylosis spondylitis, which have unknown causes, insidious onset, and progressive development, resulting in irreversible spinal joint deformities and stiffness at advanced stages and high disability rates. As one of the traditional Chinese health and fitness qigong exercises, Taijiquan has the effect of strengthening the waist and kidneys, sparing the tendons and softening the body, unblocking the meridians, unblocking the qi and blood, and strengthening the body and increasing intelligence. To this end, this paper proposes a data enhancement method that uses the spine line as the control curve to deform the human contours, drawing on the moving least squares deformation method to fit a variety of scoliosis cases and increase the diversity of the dataset, thereby improving the generalisation capability of the network. The colour map and depth map information are then integrated using a four-channel method and a dual feature extraction network structure to improve the accuracy of segmentation.

Development of a Deep Convolutional Neural Network for the Prediction of Pavement Roughness from 3D Images

Current roughness prediction models require extensive input data including pavement distress, climatic, and traffic data, which may be difficult to collect. In addition, these models have geographical limitations; therefore, a significant bias is expected if these models are used without recalibration. Pavement digital images can reflect both surface distresses and other surface irregularities that may affect pavement roughness conditions. In this study, convolutional neural networks (CNNs) were used to classify pavement sections into different roughness categories and to estimate International Roughness Index (IRI) values using pavement surface images. Furthermore, the effectiveness of artificial neural network (ANN) and multinomial logistic (MNL) regression models to categorize pavement sections into different roughness conditions was investigated. A pretrained CNN model was trained and validated using 850 three-dimensional (3D) pavement surface images, which were extracted from the Louisiana DOT and Development (LaDOTD) pavement management system (PMS) inventory. In addition, 1,142 test observations including IRI measurements and distress data were used to develop the ANN-based pattern recognition and MNL models. The developed CNN model outperformed the ANN and MNL models with an accuracy of 93.4% in the training stage. In addition, the CNN model predicted IRI values with a coefficient of determination (R2) of 0.985 and an average error of 5.9%.

Sugar beet root rot loss: ANN and Regression models

Root rot diseases of sugar beet resulted by Rhizoctonia solani, Pythium aphanidermatum can arouse significant losses in sugar beet crops. Simple and accurate statistical techniques are essential to predict sugar beet yield loss as an industrial crop for agricultural management specialists. The objectives of the present study were to compare: (1) the prediction capabilities of different statistical models at several regions and cultivars, (2) the effectiveness of artificial neural network (ANN) models to predict root rot yield loss, and (3) the effectiveness of simple and multiple linear regression models compare to ANN models. In order to measure crop loss, during two consecutive growing seasons, eleven different sugar beet cultivars were planted in the infected fields in Qazvin, Torbate Jam, Urmia and Kermanshah regions in Iran. To calculate the crop loss, the correlation between Disease incidence (DI) and the important components was considered. The highest correlation (P ≤ 0.01) was observed between DI and Root yield (RY), so we developed different yield loss models. Consequently, the best-fitted simple regression model obtained for RC01 and Urmia with 94.51 % and 90.44 % R-squared, respectively. Then the proper parameters specified in the regression models considered to develop an ANN model using Matlab software to predict the yield loss. Interestingly, obtained ANN models resulted in R2, RMSE of 87.07 %, 5.01 and the lowest AIC vs. 86.78 % and 6.033 for final linear regression, respectively. Our results suggest that, based on the simplicity of the model, simple linear regression is the best model although the artificial neural network models are more accurate for prediction.

Fractional differentiation and its use in machine learning

This article covers the implementation of fractional (non-integer order) differentiation on real data of four datasets based on stock prices of main international stock indexes: WIG 20, S&P 500, DAX and Nikkei 225. This concept has been proposed by Lopez de Prado [5] to find the most appropriate balance between zero differentiation and fully differentiated time series. The aim is making time series stationary while keeping its memory and predictive power. In addition, this paper compares fractional and classical differentiation in terms of the effectiveness of artificial neural networks. Root mean square error (RMSE) and mean absolute error (MAE) are employed in this comparison. Our investigations have determined the conclusion that fractional differentiation plays an important role and leads to more accurate predictions in case of ANN.

A Damage Detection Method Using Neural Network Optimized by Multiple Particle Collision Algorithm

A critical task of structural health monitoring is damage detection and localization. Lamb wave propagation methods have been successfully applied for damage identification in plate-like structures. However, Lamb wave processing is still a challenging task due to its multimodal and dispersive characteristics. To address this issue, data-driven machine learning approaches as artificial neural network (ANN) have been proposed. However, the effectiveness of ANN can be improved based on its architecture and the learning strategy employed to train it. The present paper proposes a Multiple Particle Collision Algorithm (MPCA) to design an optimum ANN architecture to detect and locate damages in plate-like structures. For the first time in the literature, the MPCA is applied to find damages in plate-like structures. The present work uses one piezoelectric transducer to generate Lamb wave signals on an aluminum plate structure and a linear array of four transducers to capture the scattered signals. The continuous wavelet transform (CWT) processes the captured signals to estimate the time-of-flight (ToF) that is the ANN inputs. The ANN output is the damage spatial coordinates. In addition to MPCA optimization, this paper uses a quantitative entropy-based criterion to find the best mother wavelet and the scale values. The presented experimental results show that MPCA is capable of finding a simple ANN architecture with good generalization performance in the proposed damage localization application. The proposed method is compared with the 1-dimensional convolutional neural network (1D-CNN). A discussion about the advantages and limitations of the proposed method is presented.

Artificial neural networks and regression analysis for volume estimation in native species

ABSTRACT Modeling is an important tool to estimate forest production in planted areas. Although this issue has been studied worldwide, knowledge regarding volume measurement in specific locations such as Northeast Brazil is still scarce. The present study aimed to evaluated the effectiveness of artificial neural networks (ANNs) and regression analysis in estimating the timber volume of homogeneous stands of Anadantera macrocarpa, Genipa americana, and Mimosa casalpinifolia, in order to better predict the growth and production of these species. Both methods were suitable for estimating the individual volume in 7-year-old stands with different spacing. The Spurr regression model showed better statistical results and dispersion of unbiased errors for Anadantera macrocarpa and Genipa americana, whereas the Shumacher-Hall model provided more accurate volume estimates for Mimosa caesalpinifolia. The ANNs calibrated with two neurons in the middle layer exhibited the best fit for all three species. As such, artificial neural networks can be recommended to estimate the individual volumes of the species analyzed in the study area.

Effectiveness of Artificial Neural Networks in Hedging against WTI Crude Oil Price Risk

Despite the growing share of renewable energy sources, most of the world energy supply is still based on hydrocarbons and the vast majority of world transport is fuelled by oil products. Thus, the profitability of many companies may depend on the effective management of oil price risk. In this article, we analysed the effectiveness of artificial neural networks in hedging against the risk of WTI crude oil prices increase. This was reformulated from a regressive problem to a classification problem. The effectiveness of our approach, using artificial neural networks to classify observations, was verified for over ten years of WTI futures quotes, starting from 2009. The data analysis presented in this paper confirmed that the buyer of a call option was more often likely to incur a loss as a result of its purchase than make a profit after the final payoff from the call option. The results of the conducted research confirm that neural networks can be an effective form of protection against the risk of price fluctuations. The effectiveness of a network’s operation depends on the choice of assessment indicators, but analyses show that the networks which, for the indicator that was selected, gave the best results for the training set, also resulted in positive rates of return for the test set. Significantly, we also showed interdependence between seemingly unrelated indicators: percentage of the best possible results achieved in the analysed period of time by the proposed method and percentage of all available call options that were purchased based on the results from the networks that were used.

Artificial Neural Network- (ANN-) Based Proxy Model for Fast Performances’ Forecast and Inverse Schedule Design of Steam-Flooding Reservoirs

Steam flooding is one of the most effective and mature technology in heavy oil development. In this paper, a numerical simulation technology of steam flooding reservoir based on the finite volume method is firstly established. Combined with the phase change of steam phase, the fully implicit solution for steam flooding is carried out by using adaptive-time-step Newton iteration method. The Kriging method is used for stochastically to generate 4250 geological model samples by considering reservoir heterogeneity, and corresponding production schedule parameters are randomly given; then, these reservoir model samples are handled by the numerical simulation technology to obtain corresponding dynamic production data, which constitute the data for artificial neural network (ANN) training. By using the highly nonlinear global effect of artificial neural network and its powerful self-adaptive and self-learning functions, the forward-looking and inverse design ANN models of steam-flooding reservoirs are established, which provides a new method for rapid prediction of steam-flooding production performance and production schedule parameter design. In 4250 samples, the error of the forward-looking model is basically less than 0.1%, and the error of the inverse design model is generally less than 15%. It fully shows that the ANN models developed in this paper can quickly and effectively predict oil production and design production parameters and have an important guiding role in the implementation of the steam flooding technology. Finally, the forward-looking ANN model is applied to efficiently analyze the influencing factors of steam flooding process, and uncertainty analysis of the inverse design ANN model is conducted by Monte Carlo Simulation to illustrate its robustness. Besides, this paper may provide a reference for the application of neural network models to underground oil and gas reservoir, which is a typical invisible black box.

Investigate the Effect of Artificial Neural Network Parameters to Improve Fault Distance and Impedance Accuracy

This paper investigates the effect of artificial neural network (ANN) parameters against the ANN accuracy on cable fault location. The investigation is conducted through the fault impedance and distance estimations during the occurrence of high impedance fault (HIF) in the distribution system. The measured three-phase voltage and current signals are utilized and fed into the ANN to estimate the fault impedance and distance. The accuracy of the estimated fault impedance and distance is evaluated with respect to the variation of ANN parameters. Based on the analysis, it shows that more accurate results can be obtained by utilizing the optimal value of ANN parameters.

Prediction of diabetes and hypertension using multi-layer perceptron neural networks

Background: Diabetes and hypertension are two of the commonest diseases in the world. As they unfavorably affect people of different age groups, they have become a cause of concern and must be predicted and diagnosed well in advance. Objective: This research aims to determine the effectiveness of artificial neural networks (ANNs) in predicting diabetes and blood pressure diseases and to point out the factors which have a high impact on these diseases. Sample: This work used two online datasets which consist of data collected from 768 individuals. We applied neural network algorithms to predict if the individuals have those two diseases based on some factors. Diabetes prediction is based on five factors: age, weight, fat-ratio, glucose, and insulin, while blood pressure prediction is based on six factors: age, weight, fat-ratio, blood pressure, alcohol, and smoking. Method: A model based on the Multi-Layer Perceptron Neural Network (MLP) was implemented. The inputs of the network were the factors for each disease, while the output was the prediction of the disease’s occurrence. The model performance was compared with other classifiers such as Support Vector Machine (SVM) and K-Nearest Neighbors (KNN). We used performance metrics measures to assess the accuracy and performance of MLP. Also, a tool was implemented to help diagnose the diseases and to understand the results. Result: The model predicted the two diseases with correct classification rate (CCR) of 77.6% for diabetes and 68.7% for hypertension. The results indicate that MLP correctly predicts the probability of being diseased or not, and the performance can be significantly increased compared with both SVM and KNN. This shows MLPs effectiveness in early disease prediction.

Improving Accuracy and Generalization Performance of Small-Size Recurrent Neural Networks Applied to Short-Term Load Forecasting

The load forecasting of a coal mining enterprise is a complicated problem due to the irregular technological process of mining. It is necessary to apply models that can distinguish both cyclic components and complex rules in the energy consumption data that reflect the highly volatile technological process. For such tasks, Artificial Neural Networks demonstrate advanced performance. In recent years, the effectiveness of Artificial Neural Networks has been significantly improved thanks to new state-of-the-art architectures, training methods and approaches to reduce overfitting. In this paper, the Recurrent Neural Network architecture with a small-size model was applied to the short-term load forecasting of a coal mining enterprise. A single recurrent model was developed and trained for the entire four-year operational period of the enterprise, with significant changes in the energy consumption pattern during the period. This task was challenging since it required high-level generalization performance from the model. It was shown that the accuracy and generalization properties of small-size recurrent models can be significantly improved by the proper selection of the hyper-parameters and training method. The effectiveness of the proposed approach was validated using a real-case dataset.

The effectiveness of artificial neural networks applied to analytical procedures using high level data: a simulation analysis

Purpose Auditors are required to perform analytical procedures during the planning and concluding phases of the audit. Such procedures typically use data aggregated at a high level. The authors investigate whether artificial neural networks, a more sophisticated technique for analytical review than typically used by auditors, may be effective when using high level data. Design/methodology/approach Data from companies operating in the dairy industry were used to train an artificial neural network. Data with and without material seeded errors were used to test alternative techniques. Findings Results suggest that the artificial neural network approach was not significantly more effective (taking into account both Type I and II errors) than traditional ratio and regression analysis, and none of the three approaches provided more overall effectiveness than a purely random procedure. However, the artificial neural network approach did yield considerably fewer Type II errors than the other methods, which suggests artificial neural networks could be a candidate to improve the performance of analytical procedures in circumstances where Type II error rates are the primary concern of the auditor. Originality/value The authors extend the work of Coakley and Brown (1983) by investigating the application of artificial neural networks as an analytical procedure using aggregated data. Furthermore, the authors examine multiple companies from one industry and supplement financial information with both exogenous industry and macro-economic data.

Effectiveness of Artificial Neural Network in promoting Green Supply Chain- A study conducted on select company in Bangalore

Effectiveness of Artificial Neural Network in promoting Green Supply Chain- A study conducted on select company in Bangalore

Overfitting effect of artificial neural network based nonlinear equalizer: from mathematical origin to transmission evolution

Overfitting effect of artificial neural network (ANN) based nonlinear equalizer (NLE) leads to a trap of bit error ratio (BER) overestimation in optical fiber communication system, especially when the performance is evaluated by the commonly-used pseudo-random binary sequence (PRBS). First, we mathematically investigate the PRBS generation and Gray code mapping rules, in comparison with the use of Mersenne Twister random sequence (MTRS). Under the condition of a symbol erasure channel, we identify that ANN can recognize both the PRBS generation and symbol mapping rules, by increasing the weights of NLE at specific positions, whereas the MTRS is currently safe owing to the limited input length of current ANN based NLE. Then, we design four channel models of fiber optical transmission to experimentally examine various impairments on the evolution of overfitting effect. When both the additive white Gaussian noise (AWGN) channel and the bandwidth limited channel are considered, the mitigation of overfitting becomes possible by the use of pruned PRBS (P-PRBS) training set with removing the generation and mapping rules determined input symbols. However, as for both the chromatic dispersion (CD) uncompensated channel and the CD managed channel, the overfitting effect becomes serious, because both CD and fiber nonlinearity induced inter-symbol interference (ISI) is beneficial for ANN to identify the PRBS symbol rules. Finally, possible solutions to mitigate the overfitting effect are summarized.

Developing an Intelligent System for Prediction of Optimal Dose of Warfarin in Iranian Adult Patients with Artificial Heart Valve

Introduction: Artificial intelligence (AI) research within medicine is growing rapidly. AI is poised to transform medical practice. AI has been studied in several areas of healthcare and medical practice, including diagnosing, treating and caring of patients. Warfarin is one of the most commonly prescribed oral anticoagulant. Among all anticoagulants, warfarin has long been listed among the top ten drugs causing adverse drug events. Due to narrow therapeutic range and significant side effects, warfarin dosage determination becomes a challenging task in clinical practice. The purpose of this study was to determine exact dose of warfarin needed for patients with artificial heart valve using artificial neural networks (ANN).Development: To achieved the best model, some multi-layer perceptron ANNs were constructed with different structures. The dataset used included 846 patients who had been referred to the PT clinic in Tehran heart center in the second six months of the year 2013. Finally, the best structure of ANN for warfarin dose was investigated and used for prediction system developments. In this paper the implementation of ANNs and proposed system in MatLab environment are described.Application: The effectiveness of ANNs were evaluated in terms of classification performance using 10fold cross-validation procedure and the results showed that the best model is a network that has 7 neurons in its hidden layer with an average absolute error of 0.1, disturbance rate of 0.33 and regression of 0.87. Conclusion: The achieved results reveal that ANN-based system is a suitable tool for warfarin dose prediction in Iranian patients with an artificial heartvalve. However, no system can be guaranteed to achieve 100% accuracy, but using such methods can reduce medical errors and thereby improve health care and patient safety.

ARTIFICIAL NEURAL NETWORKS AND MIXED-EFFECTS MODELING TO DESCRIBE THE STEM PROFILE OF Pinus taeda L.

The aim of this study was to compare the effectiveness of artificial neural networks (ANNs) and mixed-effects models (MEMs) in describing the stem profile of Pinus taeda L., using sample data from 246 trees. First, three taper functions of different classes were adjusted: non-segmented, segmented, and variable-form. To adjust the models, the nonlinear regression technique (nls) was used. In the best performance equation for nls-adjusted diameter estimates, the nonlinear MEM (nlme) was applied at two levels, using the age class (ci) and DBH class (cd). For this, three different study scenarios were considered, with the number of coefficients with random effects ranging from one to three in each scenario. The adjustments were made using the nls and nlme functions in R software. The selected mixed-effect equations were compared with ANNs generated in Neuro 4.0 software. The taper function models and ANNs were classified according to statistical criteria and graphical analysis of residues. The tapering equation of Bi (2000) presented better performance for diameter estimates than the non-segmented and segmented equations. Application of the nlme technique in the Bi (2000) equation increased the accuracy of the diameter estimates for Pinus taeda, in relation to the adjustment using the nls technique. In the comparison of ANNs with the variations of the Bi equation of mixed-effects, the networks performed better, indicated in the description of the P. taeda profile.

Permeability prediction and diagenesis in tight carbonates using machine learning techniques

Abstract Machine learning techniques have found their way into many problems in geoscience but have not been used significantly in the analysis of tight rocks. We present a case study testing the effectiveness of artificial neural networks and genetic algorithms for the prediction of permeability in tight carbonate rocks. The dataset consists of 130 core plugs from the Portland Formation in southern England, all of which have measurements of Klinkenberg-corrected permeability, helium porosity, characteristic pore throat diameter, and formation resistivity. Permeability has been predicted using genetic algorithms and artificial neural networks, as well as seven conventional ‘benchmark’ models with which the machine learning techniques have been compared. The genetic algorithm technique has provided a new empirical equation that fits the measured permeability better than any of the seven conventional benchmark models. However, the artificial neural network technique provided the best overall prediction method, quantified by the lowest root-mean-square error (RMSE) and highest coefficient of determination value (R2). The lowest RMSE from the conventional permeability equations was from the RGPZ equation, which predicted the test dataset with an RMSE of 0.458, while the highest RMSE came from the Berg equation, with an RMSE of 2.368. By comparison, the RMSE for the genetic algorithm and artificial neural network methods were 0.433 and 0.38, respectively. We attribute the better performance of machine learning techniques over conventional approaches to their enhanced capability to model the connectivity of pore microstructures caused by codependent and competing diagenetic processes. We also provide a qualitative model for the poroperm characteristics of tight carbonate rocks modified by each of eight diagenetic processes. We conclude that, for tight carbonate reservoirs, both machine learning techniques predict permeability more reliably and more accurately than conventional models and may be capable of distinguishing quantitatively between pore microstructures caused by different diagenetic processes.

Detection of specific features in the functioning of a system for the anti-corrosion protection of underground pipelines at oil and gas enterprises using neural networks

The information was reviewed to orderly arrange theoretical provisions and to devise practical recommendations for the diagnostic examination of a system for the anti-corrosion protection of metal oil and gas pipelines. A set of informative parameters for modeling functional relations and determining polarization potential in the system underground metal structure – cathodic protection was formed. It was proposed to apply the algorithm of prediction of corrosive current, the approach of non-linear programming, as well as the neural network, including the corresponding methods of learning, for a pipeline section, taking into account the polarization potential on the outer surface. The test set to evaluate the effectiveness of a neural network was formed. The above-mentioned information is essential for the improvement of the equipment of distant control of metal structures of oil and gas enterprises, that is, the procedures for correct measuring and evaluating direct and alternating voltages, as well as polarization potential in a pipeline. The methods and algorithms of neural networks, which are used to control the anticorrosive protection of pipelines, were explored. The study of the effectiveness of artificial neural networks, specifically, a two-layer network of direct propagation with the function of prediction of the resource of metal pipes. Using the polarization potential, we detected the capability of neural networks to perform inaccessible for conventional mathematics operations of processing, comparison, classification of images, capability of self-learning and self-organization relative to pipelines. The qualimetric quality criterion for a pipeline section, taking into consideration the optimal range of polarization potential was improved. We developed the method for prediction of the polarization potential of a cathodic protection plant and transitional specific resistance of the insulating coating on the surface of an metal structure using a neural network. Taking into consideration the results of analysis of polarization potential and transitional specific resistance, the methodology of formation of information support for procedures of degradation of anticorrosive dielectric coating and metal on the outer surface of an metal structure, as well as for predicting its resource, was designed

Effectiveness of Artificial Neural Networks in Solving Financial Time Series

This research aims to study and analyze which type of Artificial Neural Network (ANN) is more efficient and suitable in handling non-homogenous variance for financial series. Apart from addressing the behavior and efficiency of ANN, the paper also aims to present an advanced methodology for ANN, as a replacement of GARCH and ARCH models in crisis management decision makers. The application part was applied to the Egyptian exchange market, to study the local currency exchange rate volatility (1/1/2009-4/6/2013) in order to develop a model describing those changes in the exchange rate. The research concludes that the best network type to represent such financial series is the Back Propagation. Moreover, comparing the result with general regression and probabilistic networks rendered the later two inefficient at handling such series.