Integrating piecewise linear representation and weighted support vector machine for stock trading signal prediction

Piecewise linear representation (PLR) and back-propagation artificial neural network (BPN) have been integrated for the stock trading signal prediction recently (PLR–BPN). However, there are some disadvantages in avoiding over-fitting, trapping in local minimum and choosing the threshold of the trading decision. Since support vector machine (SVM) has a good way to avoid over-fitting and trapping in local minimum, we integrate PLR and weighted SVM (WSVM) to forecast the stock trading signals (PLR–WSVM). The new characteristics of PLR–WSVM are as follows: (1) the turning points obtained from PLR are set by different weights according to the change rate of the closing price between the current turning point and the next one, in which the weight reflects the relative importance of each turning point; (2) the prediction of stock trading signal is formulated as a weighted four-class classification problem, in which it does not need to determine the threshold of trading decision; (3) WSVM is used to model the relationship between the trading signal and the input variables, which improves the generalization performance of prediction model; (4) the history dataset is divided into some overlapping training–testing sets rather than training–validation–testing, which not only makes use of data fully but also reduces the time variability of data; and (5) some new technical indicators representing investors’ sentiment are added to the input variables, which improves the prediction performance. The comparative experiments among PLR–WSVM, PLR–BPN and buy-and-hold strategy (BHS) on 20 shares from Shanghai Stock Exchange in China show that the prediction accuracy and profitability of PLR–WSVM are all the best, which indicates PLR–WSVM is effective and can be used in the stock trading signal prediction.     

Using support vector machine with a hybrid feature selection method to the stock trend prediction

In this paper, we developed a prediction model based on support vector machine (SVM) with a hybrid feature selection method to predict the trend of stock markets. This proposed hybrid feature selection method, named F-score and Supported Sequential Forward Search (F_SSFS), combines the advantages of filter methods and wrapper methods to select the optimal feature subset from original feature set. To evaluate the prediction accuracy of this SVM-based model combined with F_SSFS, we compare its performance with back-propagation neural network (BPNN) along with three commonly used feature selection methods including Information gain, Symmetrical uncertainty, and Correlation-based feature selection via paired t-test. The grid-search technique using 5-fold cross-validation is used to find out the best parameter value of kernel function of SVM. In this study, we show that SVM outperforms BPN to the problem of stock trend prediction. In addition, our experimental results show that the proposed SVM-based model combined with F_SSFS has the highest level of accuracies and generalization performance in comparison with the other three feature selection methods. With these results, we claim that SVM combined with F_SSFS can serve as a promising addition to the existing stock trend prediction methods.     

Bankruptcy prediction using support vector machine with optimal choice of kernel function parameters

Bankruptcy prediction has drawn a lot of research interests in previous literature, and recent studies have shown that machine learning techniques achieved better performance than traditional statistical ones. This paper applies support vector machines (SVMs) to the bankruptcy prediction problem in an attempt to suggest a new model with better explanatory power and stability. To serve this purpose, we use a grid-search technique using 5-fold cross-validation to find out the optimal parameter values of kernel function of SVM. In addition, to evaluate the prediction accuracy of SVM, we compare its performance with those of multiple discriminant analysis (MDA), logistic regression analysis (Logit), and three-layer fully connected back-propagation neural networks (BPNs). The experiment results show that SVM outperforms the other methods.     

支持向量机理论与基于规划的神经网络学习算法

近年来支持向量机（SVM）理论得到国外学者高度的重视,普遍认为这是神经网络学习的新研究方向,近来也开始得到国内学者的注意。该文将研究SVM理论与神经网络的规划算法的关系,首先指出,Vapnik的基于SVM的算法与该文作者1994年提出的神经网络的基于规划的算法是等价的,即在样本集是线性可分的情况下,二者求得的均是最大边缘（maximal margin)解。不同的是,前者（通常用拉格郎日乘子法）求解的复杂性将随规模呈指数增长,而后者的复杂性是规模的多项式函数。其次,作者将规划算法化为求一点到某一凸集上的投影,利用这个几何的直观,给出一个构造性的迭代求解算法－－“单纯形迭代算法”。新算法有很强的几何直观性,这个直观性将加深对神经网络（线性可分情况下）学习的理解,并由此导出一个样本集是线性可分的充分必要条件。另外,新算法对知识扩充问题,给出一个非常方便的增量学习算法。最后指出,“将一些必须满足的条件,化成问题的约束条件,将网络的某一性能,作为目标函数,将网络的学习问题化为某种规划问题来求解”的原则,将是研究神经网络学习问题的一个十分有效的办法。     

A threshold-varying artificial neural network approach for classification and its application to bankruptcy prediction problem

We propose a threshold-varying artificial neural network (TV-ANN) approach for solving the binary classification problem. Using a set of simulated and real-world data set for bankruptcy prediction, we illustrate that the proposed TV-ANN fares well, both for training and holdout samples, when compared to the traditional backpropagation artificial neural network (ANN) and the statistical linear discriminant analysis. The performance comparisons of TV-ANN with a genetic algorithm-based ANN and a classification tree approach C4.5 resulted in mixed results.     

Enhanced default risk models with SVM+

Default risk models have lately raised a great interest due to the recent world economic crisis. In spite of many advanced techniques that have extensively been proposed, no comprehensive method incorporating a holistic perspective has hitherto been considered. Thus, the existing models for bankruptcy prediction lack the whole coverage of contextual knowledge which may prevent the decision makers such as investors and financial analysts to take the right decisions. Recently, SVM+ provides a formal way to incorporate additional information (not only training data) onto the learning models improving generalization. In financial settings examples of such non-financial (though relevant) information are marketing reports, competitors landscape, economic environment, customers screening, industry trends, etc. By exploiting additional information able to improve classical inductive learning we propose a prediction model where data is naturally separated into several structured groups clustered by the size and annual turnover of the firms. Experimental results in the setting of a heterogeneous data set of French companies demonstrated that the proposed default risk model showed better predictability performance than the baseline SVM and multi-task learning with SVM.     

基于快速SVM的大规模网络流量分类方法

支持向量机方法具有良好的分类准确率、稳定性与泛化性,在网络流量分类领域已有初步应用,但在面对大规模网络流量分类问题时却存在计算复杂度高、分类器训练速度慢的缺陷。为此,提出一种基于比特压缩的快速SVM方法,利用比特压缩算法对初始训练样本集进行聚合与压缩,建立具有权重信息的新样本集,在损失尽量少原始样本信息的前提下缩减样本集规模,进一步利用基于权重的SVM算法训练流量分类器。通过大规模样本集流量分类实验对比,快速SVM方法能在损失较少分类准确率的情况下,较大程度地缩减流量分类器的训练时间以及未知样本的预测时间,同时,在无过度压缩前提下,其分类准确率优于同等压缩比例下的随机取样SVM方法。本方法在保留SVM方法较好分类稳定性与泛化性能的同时,有效提升了其应对大规模流量分类问题的能力。     

Derivative and GA-based methods in metamodeling of back-propagation neural networks for constrained approximate optimization

Artificial neural networks (ANN) have been extensively used as global approximation tools in the context of approximate optimization. ANN traditionally minimizes the absolute difference between target outputs and approximate outputs thereby resulting in approximate optimal solutions being sometimes actually infeasible when it is used as a metamodel for inequality constraint functions. The paper explores the development of the efficient back-propagation neural network (BPN)-based metamodel that ensures the constraint feasibility of approximate optimal solution. The BPN architecture is optimized via two approaches of both derivative-based method and genetic algorithm (GA) to determine interconnection weights between layers in the network. The verification of the proposed approach is examined by adopting a standard ten-bar truss problem. Finally, a GA-based approximate optimization of suspension with an optical flying head is conducted to enhance the shock resistance capability in addition to dynamic characteristics.     

Bankruptcy prediction using SVM models with a new approach to combine features selection and parameter optimisation

Due to the economic significance of bankruptcy prediction of companies for financial institutions, investors and governments, many quantitative methods have been used to develop effective prediction models. Support vector machine (SVM), a powerful classification method, has been used for this task; however, the performance of SVM is sensitive to model form, parameter setting and features selection. In this study, a new approach based on direct search and features ranking technology is proposed to optimise features selection and parameter setting for 1-norm and least-squares SVM models for bankruptcy prediction. This approach is also compared to the SVM models with parameter optimisation and features selection by the popular genetic algorithm technique. The experimental results on a data set with 2010 instances show that the proposed models are good alternatives for bankruptcy prediction.     

A conservative method of wavelet neural network based meta-modeling in constrained approximate optimization

The paper aims at the development of the wavelet neural network (WNN) based conservative meta-model that satisfies the constraint feasibility of approximate optimal solution. The WNN based constraint-feasible meta-model is formulated via exterior penalty method to optimally determine interconnection weights and dilation and translation coefficients in the network. Using Ackley’s path function, the approximation performance of WNN is first tested in comparison with BPN. The proposed approach of constraint feasibility is then verified through a ten-bar planar truss problem. For constrained approximate optimization, the structural design of a composite rotor blade is explored to support the proposed strategies.     

Predictions of typhoon storm surge in Taiwan using artificial neural networks

Accurate predictions of storm surge and surge deviation are essential for industrial activities in coastal areas. Usually numerical hydrodynamic models or empirical methods are used to estimate the storm surge. This paper proposes an alternative back-propagation neural network (BPN) approach to forecast the storm surge and surge deviation. The prediction of storm surge from a previous typhoon is used as a training set to form predictions for the next event. Wind velocity, wind direction, atmospheric pressure and astronomical tide were selected as inputs in the neural network. The observations obtained during three typhoons from four stations in Taiwan were used to illustrate performance of the BPN model. Comparisons with numerical methods indicate that the storm surge and surge deviation can be efficiently predicted using BPN.     

Asymmetric bagging and random subspace for support vector machines-based relevance feedback in image retrieval

Relevance feedback schemes based on support vector machines (SVM) have been widely used in content-based image retrieval (CBIR). However, the performance of SVM-based relevance feedback is often poor when the number of labeled positive feedback samples is small. This is mainly due to three reasons: 1) an SVM classifier is unstable on a small-sized training set, 2) SVM's optimal hyperplane may be biased when the positive feedback samples are much less than the negative feedback samples, and 3) overfitting happens because the number of feature dimensions is much higher than the size of the training set. In this paper, we develop a mechanism to overcome these problems. To address the first two problems, we propose an asymmetric bagging-based SVM (AB-SVM). For the third problem, we combine the random subspace method and SVM for relevance feedback, which is named random subspace SVM (RS-SVM). Finally, by integrating AB-SVM and RS-SVM, an asymmetric bagging and random subspace SVM (ABRS-SVM) is built to solve these three problems and further improve the relevance feedback performance.     

一种新的基于SVM和主动学习的图像检索方法

在基于内容的图像检索中,支持向量机（SVM）能够很好地解决小样本问题,而主动学习算法则可以根据学习进程主动选择最佳的样本进行学习,大幅度缩短训练时间,提高分类算法效率。为使图像检索更加快速、高效,提出一种新的基于SVM和主动学习的图像检索方法。该方法根据SVM构造分类器,通过“V”型删除法快速缩减样本集,同时通过最优选择法从缩减样本集中选取最优的样本作为训练样本,最终构造出不仅信息度大而且冗余度低的最优训练样本集,从而训练出更好的SVM分类器,得到更高的检索效率。实验结果表明,与传统的SVM主动学习的图像检索方法相比,该方法能够较大幅度提高检索性能。     

面向超大数据集的SVM近似训练算法

标准SVM学习算法运行所需的时间和空间复杂度分别为O(l~3)和O(l~2),l为训练样本的数量,因此不适用于对超大数据集进行训练.提出一种基于近似解的SVM训练算法:Approximate Vector Machine(AVM).AVM采用增量学习的策略来寻找近似最优分类超平面,并且在迭代过程中采用热启动及抽样技巧来加快训练速度.理论分析表明,该算法的计算复杂度与训练样本的数量无关,因此具有良好的时间与空间扩展性.在超大数据集上的实验结果表明,该算法在极大提高训练速度的同时,仍然保持了原始分类器的泛化性能,并且训练完毕具有较少的支持向量,因此结果分类器具有更快的分类速度.     

A novel model by evolving partially connected neural network for stock price trend forecasting

This paper proposes a novel model by evolving partially connected neural networks (EPCNNs) to predict the stock price trend using technical indicators as inputs. The proposed architecture has provided some new features different from the features of artificial neural networks: (1) connection between neurons is random; (2) there can be more than one hidden layer; (3) evolutionary algorithm is employed to improve the learning algorithm and training weights. In order to improve the expressive ability of neural networks, EPCNN utilizes random connection between neurons and more hidden layers to learn the knowledge stored within the historic time series data. The genetically evolved weights mitigate the well-known limitations of gradient descent algorithm. In addition, the activation function is defined using sin(x) function instead of sigmoid function. Three experiments were conducted which are explained as follows. In the first experiment, we compared the predicted value of the trained EPCNN model with the actual value to evaluate the prediction accuracy of the model. Second experiment studied the over fitting problem which occurred in neural network training by taking different number of neurons and layers. The third experiment compared the performance of the proposed EPCNN model with other models like BPN, TSK fuzzy system, multiple regression analysis and showed that EPCNN can provide a very accurate prediction of the stock price index for most of the data. Therefore, it is a very promising tool in forecasting of the financial time series data.     

An integrative model with subject weight based on neural network learning for bankruptcy prediction

This study proposes an integration strategy regarding how to efficiently combine the currently-in-use statistical and artificial intelligence techniques. In particular, by combining multiple discriminant analysis, logistic regression, neural networks, and decision trees induction, we introduce an integrative model with subject weight based on neural network learning for bankruptcy prediction. The strength of the proposed model stems from differentiating the weights of the source methods for each subject in the testing data set. That is, the relative weights consist of N by I matrix, where N denotes the number of subjects and I denotes the number of the source methods. The experiments using a real world financial data indicate that the proposed model can marginally increase the prediction accuracy compared to the source methods. The integration strategy can be useful for a dichotomous classification problem like bankruptcy prediction since prediction can be improved by taking advantage of existing and newly emerging techniques in the future.     

PREDICTING BANK FAILURES: A NEURAL NETWORK APPROACH

The purpose of this paper is to present a neural network approach to predicting bank failures and to compare it with existing prediction methods. The task of constructing a prediction model is cast as one of training a network with a set of bankruptcy cases. Empirical results show that neural network is a competitive method among existing ones in assessing the likelihood of bank failures, especially in reducing type I misclassification rate. Issues relating to the potential and limitations of.neural network as a modeling tool are also addressed.     

Fuzzy C-means based support vector machine for channel equalisation

This paper proposes a new classification network, the fuzzy C-means based support vector machine (FCM–SVM) and applies it to channel equalisation. In contrast to a kernel-based SVM, the FCM–SVM has a smaller number of parameters while retaining the SVM's good generalisation ability. In FCM–SVM, input training data is clustered by FCM. The output of FCM–SVM is a weighted sum of the degrees where each input data belongs to the clusters. To achieve high generalisation ability, FCM–SVM weights are learned through linear kernel based SVM. Computer simulations illustrate the performance of the suggested network, where the FCM–SVM is used as a channel equaliser. Simulations with white Gaussian and coloured Gaussian noise are performed. This paper also compares simulation results from the FCM–SVM, the Gaussian kernel based SVM and the optimal equaliser.     

样例约简支持向量机

支持向量机(support vector machine,SVM)仅利用靠近分类边界的支持向量构造最优分类超平面,但求解SVM需要整个训练集,当训练集的规模较大时,求解SVM需要占用大量的内存空间,寻优速度非常慢。针对这一问题,提出了一种称为样例约简的寻找候选支持向量的方法。在该方法中,支持向量大多靠近分类边界,可利用相容粗糙集技术选出边界域中的样例,作为候选支持向量,然后将选出的样例作为训练集来求解SVM。实验结果证实了该方法的有效性,特别是对大型数据库,该方法能有效减少存储空间和执行时间。     

深度置信网络优化模型在人才评价中的应用

针对深度置信网络(DBN)权值随机初始化容易使网络陷入局部最优的问题,引入改进的和声搜索(IHS)算法,提出基于IHS的DBN模型(IHS-DBN)。在和声搜索算法的基础上,利用全局自适应的和声音调调整方式,提升算法收敛速度和后期局部搜索能力。将DBN重构误差函数作为IHS算法的优化目标函数,通过不断迭代优化解向量为DBN寻找一组较优的初始权值进行网络训练,并在MNIST数据集上验证IHS-DBN模型的有效性。IHS-DBN模型在高校人才评价中的应用结果表明,与DBN、SVM和BP神经网络评价模型相比,IHS-DBN模型的评价准确率分别提高3.6%、7.3%和16.4%。