Artificial Neural Networks-based Model Research Articles

Application of gas chromatographic data and 2D molecular descriptors for accurate global mobility potential prediction

Mobility is a key feature affecting the environmental fate, which is of particular importance in the case of persistent organic pollutants (POPs) and emerging pollutants (EPs). In this study, the global mobility classification artificial neural networks-based models employing GC retention times (RT) and 2D molecular descriptors were constructed and validated. The high usability of RT was confirmed based on the feature selection step performed using the multivariate adaptive regression splines (MARS) tool. Although RT was found to be the most important, according to Kruskal-Wallis ANOVA analysis, it is insufficient to build a robust model, which justifies the need to expand the input layer with 2D descriptors. Therefore the following molecular descriptors: MPC10, WTPT-2, AATS8s, minaaCH, GATS7c, RotBtFrac, ATSC7v and ATSC1p, which were characterized by a high predicting potential were used to improve the classification performance. As a result of machine learning procedure ten of the most accurate neural networks were selected. The external validation showed that the final models are characterized by a high general accuracy score (85.71–96.43%). The high predicting abilities were also confirmed by the micro-averaged Matthews correlation coefficient (MAMCC) (0.73–0.88). To evaluate the applicability of the models, new retention times of selected POPs and EPs including pesticides, polycyclic aromatic hydrocarbons, pharmaceuticals, fragrances and personal care products were measured and used for mobility prediction. Further, the classifiers were used for photodegradation and chlorination products of two popular sunscreen agents, 2-ethyl-hexyl-4-methoxycinnamate and 2-ethylhexyl 4-(dimethylamino)benzoate.

Multilayer perceptron artificial neural networks-based model for credit card fraud detection

Nowadays, <span>credit card fraud has emerged as a major problem. People are becoming increasingly using credit cards to pay for their transactions, it has become more popular and essential in our lives. Fraudsters are developing new strategies and techniques over time, and it is not easy for humans to manually check out all transactions. The cost of fraudulent transactions is significant and without prevention mechanisms it is rising. Finding the best methodology to detect fraudulent transactions is a crucial asset to the industry to reduce the fraud financial loss. Artificial neural networks (ANN) technique is considered as one of the effective techniques that has proved its efficiency in detecting credit card fraud transactions with high precision and minimum cost. In this paper, we propose a multilayer perceptron (MLP) ANN-based model solution to improve the accuracy of the detection process. The performance of the methodology is measured based on the precision, sensitivity, specificity, accuracy, F-measure, area under curve (AUC) and root mean square error (RMSE). Moreover, we illustrate the performance results of these measures with a descriptive analysis. Experimental results have shown that the proposed ANN-based model is efficient and does improve the accuracy of the detection of fraudulent transactions.</span>

Stock Closing Price Prediction of ISX-listed Industrial Companies Using Artificial Neural Networks

Making stock investment decisions is a complex challenge that investors continuously face. When it comes to an uncertain future, making the wrong decision can result in massive losses. The paper aims to develop an artificial neural networks-based model predicting the closing price of top-six traded industrial ISX-listed stocks, which can guide investment decisions. The sample consisted of daily indexes ISX-released from (3/3/2019) to (31/3/2019). Matlab 2014b was used to run artificial neural networks using the nntool software. The model's performance was evaluated using Mean squared error (MSE), Root means squared error (RMSE), and R squared. Empirical results demonstrated the ability and efficiency of artificial neural networks to predict closing prices with high accuracy. As a result, we recommended employing the Artificial Neural Networks model to predict stock prices as well as relying on it to make decisions.

Retraction Note To: Artificial neural network to predict the effect of heat treatments on Vickers microhardness of low-carbon Nb microalloyed steels

In the present study, an artificial neural networks-based model (ANNs) was developed to predict the Vickers microhardness of low-carbon Nb microalloyed steels. Fourteen parameters affecting the Vickers microhardness were considered as inputs, including the austenitizing temperature, cooling rate, initial austenite grain size, different chemical compositions and Nb in solution. The network was then trained to predict the Vickers microhardness amounts as outputs. A Multilayer feed-forward back-propagation network was developed and trained using experimental data form literatures. Five low-carbon Nb microalloyed steels and one low-carbon steel without Nb were investigated. The effects of austenitizing temperature (900–1,100°C) and subsequent cooling rate (0.15–227°C/s) and initial austenite grain size (5–130 μm) on the Vickers microhardness of steels were modeled by ANNs as well. The predicted values are in very good agreement with the measured ones, indicating that the developed model is very accurate and has the great ability for predicting the Vickers microhardness.

Time Series Seasonal Analysis Based on Fuzzy Transforms

We define a new seasonal forecasting method based on fuzzy transforms. We use the best interpolating polynomial for extracting the trend of the time series and generate the inverse fuzzy transform on each seasonal subset of the universe of discourse for predicting the value of an assigned output. In the first example, we use the daily weather dataset of the municipality of Naples (Italy) starting from data collected from 2003 to 2015 making predictions on mean temperature, max temperature and min temperature, all considered daily. In the second example, we use the daily mean temperature measured at the weather station “Chiavari Caperana” in the Liguria Italian Region. We compare the results with our method, the average seasonal variation, Auto Regressive Integrated Moving Average (ARIMA) and the usual fuzzy transforms concluding that the best results are obtained under our approach in both examples. In addition, the comparison results show that, for seasonal time series that have no consistent irregular variations, the performance obtained with our method is comparable with the ones obtained using Support Vector Machine- and Artificial Neural Networks-based models.

Adjusting wavelet-based multiresolution analysis boundary conditions for long-term streamflow forecasting

We propose a novel technique for improving a long-term multi-step-ahead streamflow forecast. A model based on wavelet decomposition and a multivariate Bayesian machine learning approach is developed for forecasting the streamflow 3, 6, 9, and 12 months ahead simultaneously. The inputs of the model utilize only the past monthly streamflow records. They are decomposed into components formulated in terms of wavelet multiresolution analysis. It is shown that the model accuracy can be increased by using the wavelet boundary rule introduced in this study. A simulation study is performed to evaluate the effects of different wavelet boundary rules using synthetic and real streamflow data from the Yellowstone River in the Uinta Basin in Utah. The model based on the combination of wavelet and Bayesian machine learning regression techniques is compared with that of the wavelet and artificial neural networks-based model. The robustness of the models is evaluated. Copyright © 2015 John Wiley & Sons, Ltd.

Forecasting long-term bridge deterioration conditions using artificial intelligence techniques

About a quarter of 600,000 bridges in the USA are deficient. The objective of this study is to make accurate predictions of future bridge deterioration condition using artificial neural networks, and to ensure the developed models are applicable for practical use. Using bridge inspection data provided by the Connecticut Department of Transportation ConnDOT, artificial neural networks-based model is developed to modelling complex relationships between input and output to identify patterns within the data. The developed neural net input variables included bridge geometry, construction and service, while the output variables were condition ratings for the deck, superstructure, and substructure. The neural nets used in this research demonstrated an ability to produce accurate results down to a root mean square error of 10.05% on the best trial. This study shows the potential to develop a tool to predict the future condition ratings of bridges to assist agencies in bridge program planning.

A novel unified correlation model using ensemble support vector regression for prediction of flooding velocity in randomly packed towers

Traditional empirical correlations and models have found insufficient to predict the flooding velocity accurately mainly because there are many kinds of random packings which exhibit different characteristics. In this work, a novel data-driven modeling method, i.e. ensemble least squares support vector regression (ELSSVR), is proposed to construct a unified correlation for prediction of the flooding velocity for packed towers with random packings. The flooding data are first clustered into several classes by the fuzzy c-means clustering algorithm. Then, several single LSSVR models can be trained using each sub-class of samples to capture the special characteristics. Moreover, a weighted least squares approach is adopted to integrate these single LSSVR models. Consequently, the ELSSVR model can extract the feature information of flooding data effectively and improve the prediction performance. The proposed ELSSVR method is applied to construct a unified correlation for prediction of the flooding velocity in randomly packed towers. The obtained results for several kinds of random packings demonstrate that the ELSSVR-based correlation can obtain better prediction performance, compared with the traditional semi-empirical correlations and artificial neural networks-based models. Finally, a database containing the modeling information of flooding velocity in randomly packed towers of China is provided for academic research.

ANN-based prediction of ferrite fraction in continuous cooling of microalloyed steels

In the present study, an artificial neural networks-based model was developed to predict the ferrite fraction of microalloyed steels during continuous cooling. Fourteen parameters affecting the ferrite fraction were considered as inputs, including the cooling rate, initial austenite grain size, and different chemical compositions. The network was then trained to predict the ferrite fraction amounts as outputs. A multilayer feed-forward back-propagation network was developed and trained using experimental data form literatures. The predicted values are in very good agreement with the measured ones indicating that the developed model is very accurate and has the great ability for predicting the ferrite fraction.

RETRACTED ARTICLE: Artificial neural network to predict the effect of heat treatments on Vickers microhardness of low-carbon Nb microalloyed steels

In the present study, an artificial neural networks-based model (ANNs) was developed to predict the Vickers microhardness of low-carbon Nb microalloyed steels. Fourteen parameters affecting the Vickers microhardness were considered as inputs, including the austenitizing temperature, cooling rate, initial austenite grain size, different chemical compositions and Nb in solution. The network was then trained to predict the Vickers microhardness amounts as outputs. A Multilayer feed-forward back-propagation network was developed and trained using experimental data form literatures. Five low-carbon Nb microalloyed steels and one low-carbon steel without Nb were investigated. The effects of austenitizing temperature (900–1,100°C) and subsequent cooling rate (0.15–227°C/s) and initial austenite grain size (5–130 μm) on the Vickers microhardness of steels were modeled by ANNs as well. The predicted values are in very good agreement with the measured ones, indicating that the developed model is very accurate and has the great ability for predicting the Vickers microhardness.

Identification of wear and misalignment on journal bearings using artificial neural networks

Journal bearing bushing wear and shaft misalignment are the common defects observed in rotating systems. Both defects influence their dynamic behaviour, stability and life time and their detection is a very important task. This article introduces a defect diagnosis approach that uses artificial neural networks. Reynolds equation is solved by finite element method and wear/misalignment data are provided. Then, the proposed model uses values of journal bearing performance characteristics, such as eccentricity, attitude angle and film thickness, and it trains an artificial neural network so that reliable defect identification is performed. The accuracy of the proposed model is demonstrated – for different misalignment angles, wear depths and L/ D ratios – for a worn/misaligned journal bearing. The potential use of the artificial neural networks-based model for a real-time condition monitoring system is also discussed. The outcome of this study is to obtain an efficient engineering tool that would be capable of detecting wear and misalignment of journal bearings for different operating conditions using artificial neural networks.

Artificial neural network-based predictive model for bacterial growth in a simulated medium of modified-atmosphere-packed cooked meat products.

The data of Devilieghere et al. (Int. J. Food Microbiol. 1999, 46, 57--70) on bacterial growth in a simulated medium of modified-atmosphere-packed cooked meat products was processed for estimating maximum specific growth rate mu(max) and lag phase lambda of Lactobacillus sake using artificial neural networks-based model (ANNM) computation. The comparison between ANNM and response surface methodology (RSM) model showed that the accuracy of ANNM prediction was higher than that of RSM. Two-dimensional and three-dimensional plots of the response surfaces revealed that the relationships of water activity a(w), temperature T, and dissolved CO(2) concentration with mu(max) and lambda were complicated, not just linear or second-order relations. Furthermore, it was possible to compute the sensitivity of the model outputs against each input parameter by using ANNM. The results showed that mu(max) was most sensitive to a(w), T, and dissolved CO(2) in this order; whereas lambda was sensitive to T the most, followed by a(w), and dissolved CO(2) concentrations.