Performance comparison of artificial neural network models for daily rainfall prediction

With an aim to predict rainfall one-day in advance, this paper adopted different neural network models such as feed forward back propagation neural network (BPN), cascade-forward back propagation neural network (CBPN), distributed time delay neural network (DTDNN) and nonlinear autoregressive exogenous network (NARX), and compared their forecasting capabilities. The study deals with two data sets, one containing daily rainfall, temperature and humidity data of Nilgiris and the other containing only daily rainfall data from 14 rain gauge stations located in and around Coonoor (a taluk of Nilgiris). Based on the performance analysis, NARX network outperformed all the other networks. Though there is no major difference in the performances of BPN, CBPN and DTDNN, yet BPN performed considerably well confirming its prediction capabilities. Levenberg Marquardt proved to be the most effective weight updating technique when compared to different gradient descent approaches. Sensitivity analysis was instrumental in identifying the key predictors.     

支持向量机与RBF神经网络回归性能比较研究

支持向量机与RBF神经网络相比各有优缺点,通过对支持向量机与RBF神经网络的研究,从理论上分析了这两种学习机在回归预测原理上的异同,通过仿真实验对比了两者在测试集上的逼近能力及泛化能力。仿真结果表明,对于小样本集,支持向量机的逼近能力及泛化能力要优于RBF神经网络。对实际应用中回归模型的选择问题提出了建议。     

Classification of EEG signals using neural network and logistic regression

Epileptic seizures are manifestations of epilepsy. Careful analyses of the electroencephalograph (EEG) records can provide valuable insight and improved understanding of the mechanisms causing epileptic disorders. The detection of epileptiform discharges in the EEG is an important component in the diagnosis of epilepsy. As EEG signals are non-stationary, the conventional method of frequency analysis is not highly successful in diagnostic classification. This paper deals with a novel method of analysis of EEG signals using wavelet transform and classification using artificial neural network (ANN) and logistic regression (LR). Wavelet transform is particularly effective for representing various aspects of non-stationary signals such as trends, discontinuities and repeated patterns where other signal processing approaches fail or are not as effective. Through wavelet decomposition of the EEG records, transient features are accurately captured and localized in both time and frequency context. In epileptic seizure classification we used lifting-based discrete wavelet transform (LBDWT) as a preprocessing method to increase the computational speed. The proposed algorithm reduces the computational load of those algorithms that were based on classical wavelet transform (CWT). In this study, we introduce two fundamentally different approaches for designing classification models (classifiers) the traditional statistical method based on logistic regression and the emerging computationally powerful techniques based on ANN. Logistic regression as well as multilayer perceptron neural network (MLPNN) based classifiers were developed and compared in relation to their accuracy in classification of EEG signals. In these methods we used LBDWT coefficients of EEG signals as an input to classification system with two discrete outputs: epileptic seizure or non-epileptic seizure. By identifying features in the signal we want to provide an automatic system that will support a physician in the diagnosing process. By applying LBDWT in connection with MLPNN, we obtained novel and reliable classifier architecture. The comparisons between the developed classifiers were primarily based on analysis of the receiver operating characteristic (ROC) curves as well as a number of scalar performance measures pertaining to the classification. The MLPNN based classifier outperformed the LR based counterpart. Within the same group, the MLPNN based classifier was more accurate than the LR based classifier.     

Predicting the type of pregnancy using artificial neural networks and multinomial logistic regression: a comparison study

Some medical and epidemiological surveys have been designed to predict a nominal response variable with several levels. With regard to the type of pregnancy there are four possible states: wanted, unwanted by wife, unwanted by husband and unwanted by couple. In this paper, we have predicted the type of pregnancy, as well as the factors influencing it using two different models and comparing them. Regarding the type of pregnancy with several levels, we developed a multinomial logistic regression and a neural network based on the data and compared their results using three statistical indices: sensitivity, specificity and kappa coefficient. Based on these three indices, neural network proved to be a better fit for prediction on data in comparison to multinomial logistic regression. When the relations among variables are complex, one can use neural networks instead of multinomial logistic regression to predict the nominal response variables with several levels in order to gain more accurate predictions.     

A hybrid neural network model for noisy data regression

A hybrid neural network model, based on the fusion of fuzzy adaptive resonance theory (FA ART) and the general regression neural network (GRNN), is proposed in this paper. Both FA and the GRNN are incremental learning systems and are very fast in network training. The proposed hybrid model, denoted as GRNNFA, is able to retain these advantages and, at the same time, to reduce the computational requirements in calculating and storing information of the kernels. A clustering version of the GRNN is designed with data compression by FA for noise removal. An adaptive gradient-based kernel width optimization algorithm has also been devised. Convergence of the gradient descent algorithm can be accelerated by the geometric incremental growth of the updating factor. A series of experiments with four benchmark datasets have been conducted to assess and compare effectiveness of GRNNFA with other approaches. The GRNNFA model is also employed in a novel application task for predicting the evacuation time of patrons at typical karaoke centers in Hong Kong in the event of fire. The results positively demonstrate the applicability of GRNNFA in noisy data regression problems.     

Comparison of artificial neural network and regression models for estimating software development effort

Estimating the amount of effort required for developing an information system is an important project management concern. In recent years, a number of studies have used neural networks in various stages of software development. This study compares the prediction performance of multilayer perceptron and radial basis function neural networks to that of regression analysis. The results of the study indicate that when a combined third generation and fourth generation languages data set were used, the neural network produced improved performance over conventional regression analysis in terms of mean absolute percentage error.     

计算机网络安全综合评价的神经网络模型

灰色评价法、模糊综合评价等需确定隶属函数、各指标权重,明显受人为因素的影响。尝试应用神经网络技术进行网络安全的综合评价,并通过在单指标评价标准范围内随机取值方法,生成建立神经网络模型所需的训练样本、检验样本和测试样本,在遵循BP网络建模基本原则和步骤的情况下,建立了可靠、有效的网络安全综合评价模型。16个实例研究表明：提出的样本生成方法、建模过程是可靠的,并能有效地避免出现“过训练”和“过拟合”现象,建立的BP模型具有较好的泛化能力,不受人为因素的影响,各评价指标与网络安全等级之间存在明显的非线性关系,网络安全策略对网络安全的影响最大。     

用于构建人工大脑的神经网络模型的分析与评估

目前人工脑的研究还处于起步阶段,构造智能化人工脑的方法正在探索中.影响人工脑性能的关键部分在于所选用的人工神经网络,针对目前已提出的三个网络模型,即CoDi模型、TiPo模型和DePo模型,进行了评估研究.采用的评估方法是通过解决曲线跟踪问题对模型进行测试.测试结果显示DePo模型曲线跟踪取得的效果较另两个更好,TiPo模型跟CoDi模型的性能相似.人工脑的进一步研究工作将包括提出更接近生物机制的模型或工程角度更有进化能力的模型.     

Stability evaluation of dump slope using artificial neural network and multiple regression

The present study mainly investigates the effect of the residual surface stress and the applied electric voltage on the nonlinear dynamic instability of the viscoelastic piezoelectric nanoresonators under parametric excitation. In fact, great attention is given to the influence of the residual surface stress on the nonlinear instability of the system. Numerical examples are treated which show various bifurcations. By means of a bifurcation analysis, it is shown that the instability of the system can be significantly affected by considering the residual surface effect. The results also show that a discontinuous bifurcation is always accompanied by a jump. Finally, stable and unstable regions in dynamic instability of viscoelastic piezoelectric nanoplates are addressed.

     

Landslide susceptibility assessment and factor effect analysis: backpropagation artificial neural networks and their comparison with frequency ratio and bivariate logistic regression modelling

Data collection for landslide susceptibility modeling is often an inhibitive activity. This is one reason why for quite some time landslides have been described and modelled on the basis of spatially distributed values of landslide-related attributes. This paper presents landslide susceptibility analysis in the Klang Valley area, Malaysia, using back-propagation artificial neural network model. A landslide inventory map with a total of 398 landslide locations was constructed using the data from various sources. Out of 398 landslide locations, 318 (80%) of the data taken before the year 2004 was used for training the neural network model and the remaining 80 (20%) locations (post-2004 events) were used for the accuracy assessment purpose. Topographical, geological data and satellite images were collected, processed, and constructed into a spatial database using GIS and image processing. Eleven landslide occurrence related factors were selected as: slope angle, slope aspect, curvature, altitude, distance to roads, distance to rivers, lithology, distance to faults, soil type, landcover and the normalized difference vegetation index value. For calculating the weight of the relative importance of each factor to the landslide occurrence, an artificial neural network method was developed. Each thematic layer's weight was determined by the back-propagation training method and landslide susceptibility indices (LSI) were calculated using the trained back-propagation weights. To assess the factor effects, the weights were calculated three times, using all 11 factors in the first case, then recalculating after removal of those 4 factors that had the smallest weights, and thirdly after removal of the remaining 3 least influential factors. The effect of weights in landslide susceptibility was verified using the landslide location data. It is revealed that all factors have relatively positive effects on the landslide susceptibility maps in the study. The validation results showed sufficient agreement between the computed susceptibility maps and the existing data on landslide areas. The distribution of landslide susceptibility zones derived from ANN shows similar trends as those obtained by applying in GIS-based susceptibility procedures by the same authors (using the frequency ratio and logistic regression method) and indicates that ANN results are better than the earlier method. Among the three cases, the best accuracy (94%) was obtained in the case of the 7 factors weight, whereas 11 factors based weight showed the worst accuracy (91%).     

Performance study of Byzantine Agreement Protocol with artificial neural network

Since 1982, numerous Byzantine Agreement Protocols (BAPs) have been developed to solve arbitrary faults in the Byzantine Generals Problem (BGP). A novel BAP, using an artificial neural network (ANN), was proposed by Wang and Kao. It requires message exchange rounds similar to the traditional BAP and its suitability, in the context of network size, has not been investigated. In the present study, we propose to adopt Nguyen-Widrow initialization in ANN training, which modifies message communication and limits the message exchange rounds to three rounds. This modified approach is referred to as BAP-ANN. The BAP-ANN performs better than the traditional BAP, when the network size n is greater than nine. We also evaluate the message exchange matrix (MEM) constructed during the message exchange stage. For a fixed number of faulty nodes and remainder cases of (n mod 3), the study shows that the mean epoch for ANN training decreases as the network size increases, which indicates better fault tolerance.     

An optimal tumor marker group-coupled artificial neural network for diagnosis of lung cancer

BackgroundEpidemiological statistics has shown that there are approximately 1.2 million new cases of lung cancer diagnosed every year and the death rate of these patients is 17.8%. Earlier diagnosis is key to promote the five-year survival rate of these cancer patients. Some tumor markers have been found to be valuable for earlier diagnosis, but a single marker has limitation in its sensitivity and specificity of cancer diagnosis. To improve the efficiency of diagnosis, several distinct tumor marker groups are combined together using a mathematical evaluation model, called artificial neural network (ANN). Lung cancer markers have been identified to include carcinoembryonic antigen, carcinoma antigen 125, neuron specific enolase, β₂-microglobulin, gastrin, soluble interleukin-6 receptor, sialic acid, pseudouridine, nitric oxide, and some metal ions.MethodsThese tumor markers were measured through distinct experimental procedures in 50 patients with lung cancer, 40 patients with benign lung diseases, and 50 cases for a normal control group. The most valuable were selected into an optimal tumor marker group by multiple logistic regression analysis. The optimal marker group-coupled ANN model was employed as an intelligent diagnosis system.ResultsWe have presented evidence that this system is superior to a traditional statistical method, its diagnosis specificity significantly improved from 72.0% to 100.0% and its accuracy increased from 71.4% to 92.8%.ConclusionsThe ANN-based system may provide a rapid and accurate diagnosis tool for lung cancer.     

Multiplicative neuron model artificial neural network based on Gaussian activation function

Multiplicative neuron model-based artificial neural networks are one of the artificial neural network types which have been proposed recently and have produced successful forecasting results. Sigmoid activation function was used in multiplicative neuron model-based artificial neural networks in the previous studies. Although artificial neural networks which involve the use of radial basis activation function produce more successful forecasting results, Gaussian activation function has not been used for multiplicative neuron model yet. In this study, rather than using a sigmoid activation function, Gaussian activation function was used in multiplicative neuron model artificial neural network. The weights of artificial neural network and parameters of activation functions were optimized by guaranteed convergence particle swarm optimization. Two major contributions of this study are as follows: the use of Gaussian activation function in multiplicative neuron model for the first time and the optimizing of central and propagation parameters of activation function with the weights of artificial neural network in a single optimization process. The superior forecasting performance of the proposed Gaussian activation function-based multiplicative neuron model artificial neural network was proved by applying it to real-life time series.     

Comparison of regression and artificial neural network model for the prediction of springback during air bending process of interstitial free steel sheet

This paper compares the regression and neural network modeling for predicting springback of interstial free steel sheet during air bending process. In this investigation, punch travel, strain hardening exponent, punch radius, punch velocity and width of the sheet were considered as input variables and springback as response variable. It has been observed that the ANN modeling process has been able to predict the springback with higher accuracy when compared with regression model.     

基于汽车操纵动力学的神经网络驾驶员模型

作为人一车—路闭环系统的重要环节,驾驶员模型对汽车闭环系统仿真和汽车主动安全性评价都具有重要的意义.本文基于汽车操纵动力学,预瞄—跟随理论以及神经网络建立了一种驾驶员方向控制模型,即两层前馈神经网络驾驶员模型,并在此基础上建立了驾驶员—汽车闭环系统模型.对该闭环模型进行了单移线与双移线仿真试验,仿真结果与理想数据具有很好的一致性,表明该驾驶员模型是合理的,可以有效地模拟驾驶员控制汽车方向的行为特性,为进一步研究人一车一路闭环系统提供了一条可行途径.     

Rheological wall slip velocity prediction model based on artificial neural network

Wall slip is a phenomenon in which particles migrate from solid boundaries, leaving a thin liquid rich layer adjacent to a wall, which can affect the measurement of the rheological properties. Currently, analyses of wall slip are normally carried out through experimental study using a rheometer. These traditional methods are generally time consuming, as several experiment sets are usually required. The aim of this research is to develop an alternative, more efficient approach, by formulating a mathematical model able to predict the wall slip velocity with an acceptable level of accuracy. Specifically, this study investigates a Multi-Layer Perceptron Neural Network (MLP-NN) as an advanced method to predict wall slip velocity. It develops and tests several MLP-NN architectures that accommodate a range of fixed input variables including shear stress, concentration, temperature and particle sizes, with estimated wall slip velocity as the output variable. Using this method, users can perform wall slip velocity analyses by simply plugging different patterns of the proposed input variables into the recommended architecture. Our tests show an MLP-NN model with one hidden layer consisting of nine hidden neurons to be the best architecture for such purposes, producing a strong overall performance with an R² value of 0.9994 and maximum error of 28%. This research study is innovative in its use of artificial intelligence to predict wall slip velocity in rheological applications.     

An artificial neural network model for the effects of chicken manure on ground water

In the areas where broiler industry is located, poultry manure from chicken farms could be a major source of ground water pollution, and this may have extensive effects particularly when the farms use nearby ground water as their fresh water supply. Therefore the prediction the extent of this pollution, either from rigorous mathematical diffusion modeling or from the perspective of experimental data evaluation bears importance. In this work, we have investigated modeling of the effects of chicken manure on ground water by artificial neural networks. An ANN model was developed to predict the total coliform in the ground water well in poultry farms. The back-propagation algorithm was employed for training and testing the network, and the Levenberg-Marquardt algorithm was utilized for optimization. The MATLAB 7.0 environment with Neural Network Toolbox was used for coding. Given the associated input parameters such as the number of chickens, type of manure pool management and depth of well, the model estimates the possible amount of total coliform in the wells to a satisfactory degree. Therefore it is expected to be of help in future for estimating the ground water pollution resulting from chicken farms.     

Determinants of house prices in Turkey: Hedonic regression versus artificial neural network

Determinants of house prices in Turkey are examined in this paper using the 2004 Household Budget Survey Data. In property valuation and housing market research, the locational value is usually analyzed by hedonic methods that use multiple regression techniques on large data sets and require a formality based on microeconomic theory in the analyses. Because of potential non-linearity in the hedonic functions, artificial neural network (ANN) is employed in this study as an alternative method. By comparing the prediction performance between the hedonic regression and artificial neural network models, this study demonstrates that ANN can be a better alternative for prediction of the house prices in Turkey.     

基于离散灰色预测模型与人工神经网络混合智能模型的时尚销售预测

时尚销售预测对零售领域十分重要,准确的销售情况预测有助于大幅度提高最终时尚销售利润。针对目前时尚销售预测数据量有限并且数据波动大导致难以进行准确预测的问题,提出了一种结合人工神经网络（ANN）算法和离散灰色预测模型（DGM（1,1））算法的混合智能预测算法。该算法通过关联度分析得到关联度大的影响变量,在利用DGM（1,1）+ANN预测之后,引入二次残差的思想,将实际销售数据与DGM（1,1）+ANN预测结果的残差加入影响变量利用ANN进行第二次残差预测。最后通过真实的时尚销售数据验证算法预测的可行性及准确性。实验结果表明,该算法在时尚销售数据的预测中,预测平均绝对百分误差（MAPE）在25%左右,预测性能优于自回归积分滑动平均模型（ARIMA）、扩展极限学习机（EELM）、DGM（1,1）、DGM（1,1）+ANN算法,相较于以上几种算法平均预测精度大约提高8个百分点。所提混合智能算法可用于时尚销售即时预测,且能够大幅度提高销售的效益。