Comparative Analysis of Artificial Neural Network Models: Application in Bankruptcy Prediction

This study compares the predictive performance of three neural network methods, namely the learning vector quantization, the radial basis function, and the feedforward network that uses the conjugate gradient optimization algorithm, with the performance of the logistic regression and the backpropagation algorithm. All these methods are applied to a dataset of 139 matched-pairs of bankrupt and non-bankrupt US firms for the period 1983–1994. The results of this study indicate that the contemporary neural network methods applied in the present study provide superior results to those obtained from the logistic regression method and the backpropagation algorithm.     

Temporal problems solved by dynamic fuzzy network based on genetic algorithm with variable-length chromosomes

In this paper, a dynamic fuzzy network and its design based on genetic algorithm with variable-length chromosomes is proposed. First, the dynamic fuzzy network constituted from a series of dynamic fuzzy if–then rules is proposed. One characteristic of this network is its ability to deal with temporal problems. Then, the proposed genetic algorithm with variable-length chromosomes is adopted into the design process as a means of allowing the application of the network in situations where the actual desired output is unavailable. In the proposed genetic algorithm, the length of each chromosome varies with the number of rules coded in it. Using this algorithm, no pre-assignment of the number of rules in the dynamic fuzzy network is required, since it can always help to find the most suitable number of rules. All free parameters in the network, including the spatial input partition, consequent parameters and feedback connection weights, are tuned concurrently. To further promote the design performance, genetic algorithm with variable-length chromosomes and relative-based mutated reproduction operation is proposed. In this algorithm, the elite individuals are directly reproduced to the next generation only when their averaged similarity value is smaller than a similarity threshold; otherwise, the elites are mutated to the next generation. To show the efficiency of this dynamic fuzzy network designed by genetic algorithm with variable-length chromosomes and relative-based mutated reproduction operation, two temporal problems are simulated. The simulated results and comparisons with recurrent neural and fuzzy networks verify the efficacy and efficiency of the proposed approach.     

Linear combination rule in genetic algorithm for optimization of finite impulse response neural network to predict natural chaotic time series

A finite impulse response neural network, with tap delay lines after each neuron in hidden layer, is used. Genetic algorithm with arithmetic decimal crossover and Roulette selection with normal probability mutation method with linear combination rule is used for optimization of FIR neural network. The method is applied for prediction of several important and benchmarks chaotic time series such as: geomagnetic activity index natural time series and famous Mackey–Glass time series. The results of simulations shows that applying dynamic neural models for modeling of highly nonlinear chaotic systems is more satisfactory with respect to feed forward neural networks. Likewise, global optimization method such as genetic algorithm is more efficient in comparison of nonlinear gradient based optimization methods like momentum term, conjugate gradient.     

A mended hybrid learning algorithm for radial basis function neural networks to improve generalization capability

A two-step learning scheme for radial basis function neural networks, which combines the genetic algorithm (GA) with the hybrid learning algorithm (HLA), is proposed in this paper. It is compared with the methods of the GA, the recursive orthogonal least square algorithm (ROLSA) and another two-step learning scheme for RBF neural networks, which combines the K-means clustering with the HLA (K-means + HLA). Our proposed method has the best generalization performance.     

A genetic algorithm to obtain the optimal recurrent neural network

Selecting the optimal topology of a neural network for a particular application is a difficult task. In the case of recurrent neural networks, most methods only induce topologies in which their neurons are fully connected. In this paper, we present a genetic algorithm capable of obtaining not only the optimal topology of a recurrent neural network but also the least number of connections necessary. Finally, this genetic algorithm is applied to a problem of grammatical inference using neural networks, with very good results.     

A hybrid genetic algorithm–neural network strategy for simulation optimization

Simulation optimization aims at determining the best values of input parameters, while the analytical objective function and constraints are not explicitly known in terms of design variables and their values only can be estimated by complicated analysis or time-consuming simulation. In this paper, a hybrid genetic algorithm–neural network strategy (GA–NN) is proposed for such kind of optimization problems. The good approximation performance of neural network (NN) and the effective and robust evolutionary searching ability of genetic algorithm (GA) are applied in hybrid sense, where NNs are employed in predicting the objective value, and GA is adopted in searching optimal designs based on the predicted fitness values. Numerical simulation results and comparisons based on a well-known pressure vessel design problem demonstrate the feasibility and effectiveness of the framework, and much better results are achieved than some existed literature results.     

G-PNN: A genetically engineered probabilistic neural network

The probabilistic neural network (PNN) is a neural network architecture that approximates the functionality of the Bayesian classifier, the optimal classifier. Designing the optimal Bayesian classifier is infeasible in practice, since the distributions of data belonging to each class are unknown. PNN is an approximation of the Bayesian classifier by approximating these distributions using the Parzen window approach. One of the criticisms of the PNN classifier is that, at times, it uses a lot of training data for its design. Furthermore, the PNN classifier requires that the user specifies certain network parameters, called the smoothing (spread) parameters, in order to approximate the distributions of the class data, which is not an easy task. A number of approaches have been reported in the literature for addressing both of these issues (i.e., reducing the number of training data needed for the building of the PNN model and producing good values for the smoothing parameters). In this effort, genetic algorithms are used to achieve both goals at once, and some promising results are reported.     

三种生成神经网络拓扑结构的方法在股票、商品价格预测中的应用及结果比较

对比了三种不同神经网络模型的生成方式:传统神经网络生成模型,遗传算法训练神经网络模型,以及在第二种方式训练参数的基础上,再使用传统神经网络优化生成模型.论文使用上述三种方法对代表性股票和商品价格进行拟合并预测,通过预测结果准确性和稳定性的比较发现:引入遗传算法后的神经网络在样本内的拟合误差有所降低,而第三种方法在样本外有最低的预测误差和最优稳定性.     

Image compression with a dynamic autoassociative neural network

Image compression using neural networks has been attempted with some promise. Among the architectures, feedforward backpropagation networks (FFBPN) have been used in several attempts. Although it is demonstrated that using the mean quadratic error function is equivalent to applying the Karhunen-Loeve transformation method, promise still arises from directed learning possibilities, generalization abilities and performance of the network once trained. In this paper we propose an architecture and an improved training method to attempt to solve some of the shortcomings of traditional data compression systems based on feedforward neural networks trained with backpropagation—the dynamic autoassociation neural network (DANN).The successful application of neural networks to any task requires proper training of the network. In this research, this issue is taken as the main consideration in the design of DANN. We emphasize the convergence of the learning process proposed by DANN. This process provides an escape mechanism, by adding neurons in a random state, to avoid the local minima trapping seen in traditional PFBPN. Also, DANN's training algorithm constrains the error for every pattern to an allowed interval to balance the training for every pattern, thus improving the performance rates in recognition and generalization. The addition of these two mechanisms to DANN's training algorithm has the result of improving the final quality of the images processed by DANN.The results of several tasks presented to DANN-based compression are compared and contrasted with the performance of an FFBPN-based system applied to the same task. These results indicate that DANN is superior to FFBPN when applied to image compression.     

Supply chain network design: partner selection and production/distribution planning using a systematic model

In this paper, a novel multi-phase mathematical approach is presented for the design of a complex supply chain network. From the point of network design, customer demands, and for maximum overall utility, the important issues are to find suitable and quality companies, and to decide upon an appropriate production/distribution strategy. The proposed approach is based on the genetic algorithm (GA), the analytical hierarchy process (AHP), and the multi-attribute utility theory (MAUT) to satisfy simultaneously the preferences of the suppliers and the customers at each level in the network. A case study with a good quality solution is provided to confirm the efficiency and effectiveness of the proposed approach. Finally, to demonstrate the performance of the proposed approach, a comparative numerical experiment is performed by using the proposed approach and the common single-phase genetic algorithm (SGA). Empirical analysis results demonstrate that the proposed approach can outperform the SGA in partner selection and production/distribution planning for network design.     

基于ANP-RBF神经网络的化工行业绿色供应商选择

基于绿色供应链理念,提出了化工行业绿色供应商选择的特色指标,构建了化工行业绿色供应商选择的ANP-RBF神经网络模型。通过ANP确定各指标权重,再结合RBF神经网络,从训练数据中提取隐含的知识和规律,能够方便地用于新供应商的选择。该模型求解算法为增量算法,具有很好的可扩展性,从而增加了评价的动态性。算例验证结果表明,将ANP-RBF神经网络模型用于化工行业绿色供应商的选择具有较强的实用性。     

一种新的基于SVM参数优化的算法

针对单纯使用遗传算法处理大规模数据需要时间长和对计算机的内存等硬件要求较高的问题,将神经网络嵌入到遗传算法中构造出混合智能遗传算法用于SVM核函数的参数优化,数值试验结果表明该算法对SVM核参数优化是可行的、有效的,并能得到较好的SVM核参数组合和具有较高的分类准确率及较好的泛化能力.     

BP-GA混合优化策略在人力资源战略规划中的应用

采用混合优化策略训练神经网络,进而实现地区人力资源数据的时间序列预测.神经网络,尤其是应用反向传播(back propagation,简称BP)算法训练的神经网络,被广泛应用于预测中.但是BP神经网络训练速度慢、容易陷入局部极值.遗传算法(genetic algorithm,简称GA)具有很好的全局寻优性.因而提出将BP和GA结合起来的混合优化策略训练神经网络,来实现人力资源数据预测.与BP算法相比,数值计算结果表明预测精度高、速度快,为地区人力资源数据的时间序列预测研究提供了一条新的途径.     

Distributed subgradient method for multi‐agent optimization with quantized communication

This paper focuses on a distributed optimization problem associated with a time‐varying multi‐agent network with quantized communication, where each agent has local access to its convex objective function, and cooperatively minimizes a sum of convex objective functions of the agents over the network. Based on subgradient methods, we propose a distributed algorithm to solve this problem under the additional constraint that agents can only communicate quantized information through the network. We consider two kinds of quantizers and analyze the quantization effects on the convergence of the algorithm. Furthermore, we provide explicit error bounds on the convergence rates that highlight the dependence on the quantization levels. Finally, some simulation results on a l₁‐regression problem are presented to demonstrate the performance of the algorithm. Copyright © 2016 John Wiley & Sons, Ltd.     

Application of a genetic algorithm to compute the control of a complex dynamic system

A genetic algorithm is applied to compute the control of a complex dynamic system. Methods are proposed to modify the genetic algorithm so as to adapt it to variation of the objective function. Comparison with dynamic programming results is carried out. The study was supported by the Russian Foundation of Basic Research (grant No. 9607-89110) and Ministry of Science (grants 0201.03.002 and 0201.03.003). Translated from Chislennye Metody i Vychislitel'nyi Eksperiment, Moscow State University, pp. 82–96, 1998.     

三层前向人工神经网络全局最优逼近

提出了求解不等式约束非线性优化问题的群体复合形进化算法 ,提出的算法能充分利用目标函数值的信息、优化搜索过程具有较强的方向性和目标性 ,收敛速度较快 ,且是全局优化算法 ;将群体复合形进化算法应用于三层前向人工神经网络逼近 ,提出了三层前向人工神经网络全局最优逼近算法 ;将三层前向人工神经网络全局最优逼近算法应用于实例 ,表明了提出的全局最优逼近算法的有效性 .     

Wind energy prediction using a two-hidden layer neural network

The power generated by wind turbines changes rapidly because of the continuous fluctuation of wind speed and air density. As a consequence, it can be important to predict the energy production, starting from some basic input parameters. The aim of this paper is to show that a two-hidden layer neural network can represent a useful tool to carefully predict the wind energy output. By using proper experimental data (collected from three wind farm in Southern Italy) in combination with a back propagation learning algorithm, a suitable neural architecture is found, characterized by the hyperbolic tangent transfer function in the first hidden layer and the logarithmic sigmoid transfer function in the second hidden layer. Simulation results are reported, showing that the estimated wind energy values (predicted by the proposed network) are in good agreement with the experimental measured values.     

Knowledge acquisition and revision using neural networks: an application to a cross-national study of brand image perception

A three-tier knowledge management approach is proposed in the context of a cross-national study of car brand and corporate image perceptions. The approach consists of knowledge acquisition, transfer and revision using neural networks. We investigate how knowledge acquired by a neural network from one car market can be exploited and applied in another market. This transferred knowledge is subsequently revised for application in the new market. Knowledge revision is achieved by re-training the neural network. Core knowledge common to both markets is retained while some localized knowledge components are introduced during network re-training. Since the knowledge acquired by a neural network can be expressed as an accurate set of simple rules, we are able to compare the knowledge extracted from one network with the knowledge extracted from another. Comparison of the originally acquired knowledge with the revised knowledge provides us with insights into the commonalities and differences in car brand and corporate perceptions across national markets.     

On the combination of kernel principal component analysis and neural networks for process indirect control

ABSTRACT

A new adaptive kernel principal component analysis (KPCA) for non-linear discrete system control is proposed. The proposed approach can be treated as a new proposition for data pre-processing techniques. Indeed, the input vector of neural network controller is pre-processed by the KPCA method. Then, the obtained reduced neural network controller is applied in the indirect adaptive control. The influence of the input data pre-processing on the accuracy of neural network controller results is discussed by using numerical examples of the cases of time-varying parameters of single-input single-output non-linear discrete system and multi-input multi-output system. It is concluded that, using the KPCA method, a significant reduction in the control error and the identification error is obtained. The lowest mean squared error and mean absolute error are shown that the KPCA neural network with the sigmoid kernel function is the best.     

On the enrouting protocol problem under uncertainty

In this paper we present a two-stage stochastic mixed 0–1 dynamic multicommodity model and algorithm for determining the enrouting protocol in the telecommunications network under uncertainty. Given the network connectivity, node processing and buffer and arc flow capacity, the aim is to determine the outgoing arc for the information flow reaching a given node for each destination terminal node (i.e., obtaining the route to be followed by the information flow from each origin terminal node to each destination terminal node). The origin–destination (O–D) flow matrix is given by the number of information packets to be sent from the origin terminal nodes to the destination terminal nodes along a given time horizon, i.e., a call scale. The uncertainty in the O–D flow matrix is treated via a scenario tree approach. The main goal is to minimize a composite function of the expected lost information, a penalization of the deviation from the FIFO strategy on the information flow entering the network, and the expected number of nodes visited by the information packets. A mixture of an enrouting arc generation scheme and a genetic algorithm for obtaining the enrouting protocols over the scenarios is presented. The tool presented in this paper could be used for simulating the enrouting protocols to analyze the saturation of the network, but it has a time constraint for real time operation. Faster algorithms are needed to define the routing tables during the operation stage. Computational experience is reported.