Hybridizing kernel‐based fuzzy c‐means with hierarchical selective neural network ensemble model for business failure prediction

More and more ensemble models are used to forecast business failure. It is generally known that the performance of an ensemble relies heavily on the diversity between each base classifier. To achieve diversity, this study uses kernel‐based fuzzy c‐means (KFCM) to organize firm samples and designs a hierarchical selective ensemble model for business failure prediction (BFP). First, three KFCM methods—Gaussian KFCM (GFCM), polynomial KFCM (PFCM), and Hyper‐tangent KFCM (HFCM)—are employed to partition the financial data set into three data sets. A neural network (NN) is then adopted as a basis classifier for BFP, and three sets, which are derived from three KFCM methods, are used to build three classifier pools. Next, classifiers are fused by the two‐layer hierarchical selective ensemble method. In the first layer, classifiers are ranked based on their prediction accuracy. The stepwise forward selection method is employed to selectively integrate classifiers according to their accuracy. In the second layer, three selective ensembles in the first layer are integrated again to acquire the final verdict. This study employs financial data from Chinese listed companies to conduct empirical research, and makes a comparative analysis with other ensemble models and all its component models. It is the conclusion that the two‐layer hierarchical selective ensemble is good at forecasting business failure.     

Rough sets bankruptcy prediction models versus auditor signalling rates

Both international and US auditing standards require auditors to evaluate the risk of bankruptcy when planning an audit and to modify their audit report if the bankruptcy risk remains high at the conclusion of the audit. Bankruptcy prediction is a problematic issue for auditors as the development of a cause–effect relationship between attributes that may cause or be related to bankruptcy and the actual occurrence of bankruptcy is difficult. Recent research indicates that auditors only signal bankruptcy in about 50% of the cases where companies subsequently declare bankruptcy. Rough sets theory is a new approach for dealing with the problem of apparent indiscernibility between objects in a set that has had a reported bankruptcy prediction accuracy ranging from 76% to 88% in two recent studies. These accuracy levels appear to be superior to auditor signalling rates, however, the two prior rough sets studies made no direct comparisons to auditor signalling rates and either employed small sample sizes or non‐current data. This study advances research in this area by comparing rough set prediction capability with actual auditor signalling rates for a large sample of United States companies from the 1991 to 1997 time period. Prior bankruptcy prediction research was carefully reviewed to identify 11 possible predictive factors which had both significant theoretical support and were present in multiple studies. These factors were expressed as variables and data for 11 variables was then obtained for 146 bankrupt United States public companies during the years 1991–1997. This sample was then matched in terms of size and industry to 145 non‐bankrupt companies from the same time period. The overall sample of 291 companies was divided into development and validation subsamples. Rough sets theory was then used to develop two different bankruptcy prediction models, each containing four variables from the 11 possible predictive variables. The rough sets theory based models achieved 61% and 68% classification accuracy on the validation sample using a progressive classification procedure involving three classification strategies. By comparison, auditors directly signalled going concern problems via opinion modifications for only 54% of the bankrupt companies. However, the auditor signalling rate for bankrupt companies increased to 66% when other opinion modifications related to going concern issues were included. In contrast with prior rough sets theory research which suggested that rough sets theory offered significant bankruptcy predictive improvements for auditors, the rough sets models developed in this research did not provide any significant comparative advantage with regard to prediction accuracy over the actual auditors' methodologies. The current research results should be fairly robust since this rough sets theory based research employed (1) a comparison of the rough sets model results to actual auditor decisions for the same companies, (2) recent data, (3) a relatively large sample size, (4) real world bankruptcy/non‐bankruptcy frequencies to develop the variable classifications, and (5) a wide range of industries and company sizes. Copyright © 2003 John Wiley & Sons, Ltd.     

Severity Prediction of Traffic Accident Using an Artificial Neural Network

In this paper, an artificial neural network (ANN) was used to predict the injury severity of traffic accidents based on 5973 traffic accident records occurred in Abu Dhabi over a 6‐year period (from 2008 to 2013). For each accident record, 48 different attributes had been collected at the time of the accident. After data preprocessing, the data were reduced to 16 attributes and four injury severity classes. In this study, WEKA (Waikato Environment for Knowledge Analysis) data‐mining software was used to build the ANN classifier. The traffic accident data were used to build two classifiers in two different ways. The whole data set were used for training and validating the first classifier (training set), while 90% of the data were used for training the second classifier and the remaining 10% were used for testing it (testing set). The experimental results revealed that the developed ANN classifiers can predict accident severity with reasonable accuracy. The overall model prediction performance for the training and testing data were 81.6% and 74.6%, respectively. To improve the prediction accuracy of the ANN classifier, traffic accident data were split into three clusters using a k‐means algorithm. The results after clustering revealed significant improvement in the prediction accuracy of the ANN classifier, especially for the training dataset. In this work, and in order to validate the performance of the ANN model, an ordered probit model was also used as a comparative benchmark. The dependent variable (i.e. degree of injury) was transformed from ordinal to numerical (1, 2, 3, 4) for (minor, moderate, sever, death). The R tool was used to perform an ordered probit. For each accident, the ordered probit model showed how likely this accident would result in each class (minor, moderate, severe, death). The accuracy of 59.5% obtained from the ordered probit model was clearly less than the ANN accuracy value of 74.6%. Copyright © 2016 John Wiley & Sons, Ltd.     

Foreign exchange market prediction with multiple classifiers

Foreign exchange market prediction is attractive and challenging. According to the efficient market and random walk hypotheses, market prices should follow a random walk pattern and thus should not be predictable with more than about 50% accuracy. In this article, we investigate the predictability of foreign exchange spot rates of the US dollar against the British pound to show that not all periods are equally random. We used the Hurst exponent to select a period with great predictability. Parameters for generating training patterns were determined heuristically by auto‐mutual information and false nearest‐neighbor methods. Some inductive machine‐learning classifiers—artificial neural network, decision tree, k‐nearest neighbor, and naïve Bayesian classifier—were then trained with these generated patterns. Through appropriate collaboration of these models, we achieved a prediction accuracy of up to 67%. Copyright © 2009 John Wiley & Sons, Ltd.     

Incorporating textual and management factors into financial distress prediction: A comparative study of machine learning methods

Financial distress prediction (FDP) has been widely considered as a promising approach to reducing financial losses. While financial information comprises the traditional factors involved in FDP, nonfinancial factors have also been examined in recent studies. In light of this, the purpose of this study is to explore the integrated factors and multiple models that can improve the predictive performance of FDP models. This study proposes an FDP framework to reveal the financial distress features of listed Chinese companies, incorporating financial, management, and textual factors, and evaluating the prediction performance of multiple models in different time spans. To develop this framework, this study employs the wrapper-based feature selection method to extract valuable features, and then constructs multiple single classifiers, ensemble classifiers, and deep learning models in order to predict financial distress. The experiment results indicate that management and textual factors can supplement traditional financial factors in FDP, especially textual ones. This study also discovers that integrated factors collected 4 years prior to the predicted benchmark year enable a more accurate prediction, and the ensemble classifiers and deep learning models developed can achieve satisfactory FDP performance. This study makes a novel contribution as it expands the predictive factors of financial distress and provides new findings that can have important implications for providing early warning signals of financial risk.     

Application of Machine Learning Methods to Risk Assessment of Financial Statement Fraud: Evidence from China

This study presents a method of assessing financial statement fraud risk. The proposed approach comprises a system of financial and non‐financial risk factors, and a hybrid assessment method that combines machine learning methods with a rule‐based system. Experiments are performed using data from Chinese companies by four classifiers (logistic regression, back‐propagation neural network, C5.0 decision tree and support vector machine) and an ensemble of those classifiers. The proposed ensemble of classifiers outperform each of the four classifiers individually in accuracy and composite error rate. The experimental results indicate that non‐financial risk factors and a rule‐based system help decrease the error rates. The proposed approach outperforms machine learning methods in assessing the risk of financial statement fraud. Copyright © 2014 John Wiley & Sons, Ltd.     

A semiparametric method for predicting bankruptcy

Bankruptcy prediction methods based on a semiparametric logit model are proposed for simple random (prospective) and case–control (choice‐based; retrospective) data. The unknown parameters and prediction probabilities in the model are estimated by the local likelihood approach, and the resulting estimators are analyzed through their asymptotic biases and variances. The semiparametric bankruptcy prediction methods using these two types of data are shown to be essentially equivalent. Thus our proposed prediction model can be directly applied to data sampled from the two important designs. One real data example and simulations confirm that our prediction method is more powerful than alternatives, in the sense of yielding smaller out‐of‐sample error rates. Copyright © 2007 John Wiley & Sons, Ltd.     

A Robust Data‐Mining Approach to Bankruptcy Prediction

In this study, new variants of genetic programming (GP), namely gene expression programming (GEP) and multi‐expression programming (MEP), are utilized to build models for bankruptcy prediction. Generalized relationships are obtained to classify samples of 136 bankrupt and non‐bankrupt Iranian corporations based on their financial ratios. An important contribution of this paper is to identify the effective predictive financial ratios on the basis of an extensive bankruptcy prediction literature review and upon a sequential feature selection analysis. The predictive performance of the GEP and MEP forecasting methods is compared with the performance of traditional statistical methods and a generalized regression neural network. The proposed GEP and MEP models are effectively capable of classifying bankrupt and non‐bankrupt firms and outperform the models developed using other methods. Copyright © 2011 John Wiley & Sons, Ltd.     

A Time‐Simultaneous Prediction Box for a Multivariate Time Series

A sample‐based method in Kolsrud (Journal of Forecasting 2007; 26 (3): 171–188) for the construction of a time‐simultaneous prediction band for a univariate time series is extended to produce a variable‐ and time‐simultaneous prediction box for a multivariate time series. A measure of distance based on the L_∞ ‐norm is applied to a learning sample of multivariate time trajectories, which can be mean‐ and/or variance‐nonstationary. Based on the ranking of distances to the centre of the sample, a subsample of the most central multivariate trajectories is selected. A prediction box is constructed by circumscribing the subsample with a hyperrectangle. The fraction of central trajectories selected into the subsample can be calibrated by bootstrap such that the expected coverage of the box equals a prescribed nominal level. The method is related to the concept of data depth, and thence modified to increase coverage. Applications to simulated and empirical data illustrate the method, which is also compared to several other methods in the literature adapted to the multivariate setting. Copyright © 2015 John Wiley & Sons, Ltd.     

Time‐simultaneous prediction band for a time series

I propose principles and methods for the construction of a time‐simultaneous prediction band for a univariate time series. The methods are entirely based on a learning sample of time trajectories, and make no parametric assumption about its distribution. Hence, the methods are general and widely applicable. The expected coverage probability of a band can be estimated by a bootstrap procedure. The estimate is likely to be less than the nominal level. Expected lack of coverage can be compensated for by increasing the coverage in the learning sample. Applications to simulated and empirical data illustrate the methods. Copyright © 2007 John Wiley & Sons, Ltd.     

Nonparametric Quantile Regression‐Based Classifiers for Bankruptcy Forecasting

An improved classification device for bankruptcy forecasting is proposed. The proposed approach relies on mainstream classifiers whose inputs are obtained from a so‐called multinorm analysis, instead of traditional indicators such as the ROA ratio and other accounting ratios. A battery of industry norms (computed by using nonparametric quantile regressions) is obtained, and the deviations of each firm from this multinorm system are used as inputs for the classifiers. The approach is applied to predict bankruptcy on a representative sample of Spanish manufacturing firms. Results indicate that our proposal may significantly enhance predictive accuracy, both in linear and nonlinear classifiers. Copyright © 2013 John Wiley & Sons, Ltd.     

A Bayesian nonlinear support vector machine error correction model

The use of linear error correction models based on stationarity and cointegration analysis, typically estimated with least squares regression, is a common technique for financial time series prediction. In this paper, the same formulation is extended to a nonlinear error correction model using the idea of a kernel‐based implicit nonlinear mapping to a high‐dimensional feature space in which linear model formulations are specified. Practical expressions for the nonlinear regression are obtained in terms of the positive definite kernel function by solving a linear system. The nonlinear least squares support vector machine model is designed within the Bayesian evidence framework that allows us to find appropriate trade‐offs between model complexity and in‐sample model accuracy. From straightforward primal–dual reasoning, the Bayesian framework allows us to derive error bars on the prediction in a similar way as for linear models and to perform hyperparameter and input selection. Starting from the results of the linear modelling analysis, the Bayesian kernel‐based prediction is successfully applied to out‐of‐sample prediction of an aggregated equity price index for the European chemical sector. Copyright © 2006 John Wiley & Sons, Ltd.     

On the directional predictability of equity premium using machine learning techniques

This paper applies a plethora of machine learning techniques to forecast the direction of the US equity premium. Our techniques include benchmark binary probit models, classification and regression trees, along with penalized binary probit models. Our empirical analysis reveals that the sophisticated machine learning techniques significantly outperformed the benchmark binary probit forecasting models, both statistically and economically. Overall, the discriminant analysis classifiers are ranked first among all the models tested. Specifically, the high-dimensional discriminant analysis classifier ranks first in terms of statistical performance, while the quadratic discriminant analysis classifier ranks first in economic performance. The penalized likelihood binary probit models (least absolute shrinkage and selection operator, ridge, elastic net) also outperformed the benchmark binary probit models, providing significant alternatives to portfolio managers.     

An Adaptive Multiscale Ensemble Learning Paradigm for Nonstationary and Nonlinear Energy Price Time Series Forecasting

For forecasting nonstationary and nonlinear energy prices time series, a novel adaptive multiscale ensemble learning paradigm incorporating ensemble empirical mode decomposition (EEMD), particle swarm optimization (PSO) and least square support vector machines (LSSVM) with kernel function prototype is developed. Firstly, the extrema symmetry expansion EEMD, which can effectively restrain the mode mixing and end effects, is used to decompose the energy price into simple modes. Secondly, by using the fine‐to‐coarse reconstruction algorithm, the high‐frequency, low‐frequency and trend components are identified. Furthermore, autoregressive integrated moving average is applicable to predicting the high‐frequency components. LSSVM is suitable for forecasting the low‐frequency and trend components. At the same time, a universal kernel function prototype is introduced for making up the drawbacks of single kernel function, which can adaptively select the optimal kernel function type and model parameters according to the specific data using the PSO algorithm. Finally, the prediction results of all the components are aggregated into the forecasting values of energy price time series. The empirical results show that, compared with the popular prediction methods, the proposed method can significantly improve the prediction accuracy of energy prices, with high accuracy both in the level and directional predictions. Copyright © 2016 John Wiley & Sons, Ltd.     

Google's MIDAS Touch: Predicting UK Unemployment with Internet Search Data

Internet search data could be a useful source of information for policymakers when formulating decisions based on their understanding of the current economic environment. This paper builds on earlier literature via a structured value assessment of the data provided by Google Trends. This is done through two empirical exercises related to the forecasting of changes in UK unemployment. Firstly, economic intuition provides the basis for search term selection, with a resulting Google indicator tested alongside survey‐based variables in a traditional forecasting environment. Secondly, this environment is expanded into a pseudo‐time nowcasting framework which provides the backdrop for assessing the timing advantage that Google data have over surveys. The framework is underpinned by a MIDAS regression which allows, for the first time, the easy incorporation of Internet search data at its true sampling rate into a nowcast model for predicting unemployment. Copyright © 2016 John Wiley & Sons, Ltd.     

Using accounting‐based information on young firms to predict bankruptcy

This study analyzes the nonlinear relationships between accounting‐based key performance indicators and the probability that the firm in question will become bankrupt or not. The analysis focuses particularly on young firms and examines whether these nonlinear relationships are affected by a firm's age. The analysis of nonlinear relationships between various predictors of bankruptcy and their interaction effects is based on a structured additive regression model and on a comprehensive data set on German firms. The results of this analysis provide empirical evidence that a firm's age has a considerable effect on how accounting‐based key performance indicators can be used to predict the likelihood that a firm will go bankrupt. More specifically, the results show that there are differences between older firms and young firms with respect to the nonlinear effects of the equity ratio, the return on assets, and the sales growth on their probability of bankruptcy.     

Judgement in learning‐curve forecasting: a laboratory study

This study investigates whether human judgement can be of value to users of industrial learning curves, either alone or in conjunction with statistical models. In a laboratory setting, it compares the forecast accuracy of a statistical model and judgemental forecasts, contingent on three factors: the amount of data available prior to forecasting, the forecasting horizon, and the availability of a decision aid (projections from a fitted learning curve). The results indicate that human judgement was better than the curve forecasts overall. Despite their lack of field experience with learning curve use, 52 of the 79 subjects outperformed the curve on the set of 120 forecasts, based on mean absolute percentage error. Human performance was statistically superior to the model when few data points were available and when forecasting further into the future. These results indicate substantial potential for human judgement to improve predictive accuracy in the industrial learning‐curve context. Copyright © 1999 John Wiley & Sons, Ltd.     

Residual income,non‐earnings information,and information content

We extend Ohlson's (1995) model and examine the relationship between returns and residual income that incorporate analysts' earnings forecasts and other non‐earnings information variables in the balance sheet, namely default probability and agency cost of a debt covenant contract. We further divide the sample based on bankruptcy (agency) costs, earnings components and growth opportunities of a firm to explore how these factors affect the returns–residual income link. We find that the relative predictive ability for contemporaneous stock price by considering other earnings and non‐earnings information is better than that of models without non‐earnings information. If the bankruptcy (agency) cost of a firm is higher, its information role in the firm's equity valuation becomes more important and the accuracy of price prediction is therefore higher. As for non‐earnings information, if bankruptcy (agency) cost is lower, the information role becomes more relevant, and the earnings response coefficient is hence higher. Moreover, the decomposition of unexpected residual income into permanent and transitory components induces more information than that of the unexpected residual income alone. The permanent component has a larger impact than the transitory component in explaining abnormal returns. The market and industry properties and growth opportunity also have incremental explanatory power in valuation. Copyright © 2008 John Wiley & Sons, Ltd.     

A novel hybrid fine particulate matter (PM2.5) forecasting and its further application system: Case studies in China

Air pollution has received more attention from many countries and scientists due to its high threat to human health. However, air pollution prediction remains a challenging task because of its nonstationarity, randomness, and nonlinearity. In this research, a novel hybrid system is successfully developed for PM_2.5 concentration prediction and its application in health effects and economic loss assessment. First, an efficient data mining method is adopted to capture and extract the primary characteristic of PM_2.5 dataset and alleviate the noises' adverse effects. Second, Harris hawks optimization algorithm is introduced to tune the extreme learning machine model with high prediction accuracy, then the optimized extreme learning machine can be established to obtain the forecasting values of PM_2.5 series. Next, PM_2.5-related health effects and economic costs was estimated based on the predicted PM_2.5 values, the related health effects, and environmental value assessment methods. Several experiments are designed using three daily PM_2.5 datasets from Beijing, Tianjin, and Shijiazhuang. Lastly, the corresponding experimental results showed that this proposed system can not only provide early warning information for environmental management, assist in the formulation of effective measures to reduce air pollutant emissions, and prevent health problems but also help for further research and application in different fields, such as health issues due to PM_2.5 pollutant.     

Ex post and ex ante prediction of unobserved multivariate time series: a structural‐model based approach

A methodology for estimating high‐frequency values of an unobserved multivariate time series from low‐frequency values of and related information to it is presented in this paper. This is an optimal solution, in the multivariate setting, to the problem of ex post prediction, disaggregation, benchmarking or signal extraction of an unobservable stochastic process. Also, the problem of extrapolation or ex ante prediction is optimally solved and, in this context, statistical tests are developed for checking online the ocurrence of extreme values of the unobserved time series and consistency of future benchmarks with the present and past observed information. The procedure is based on structural or unobserved component models, whose assumptions and specification are validated with the data alone. Copyright © 2007 John Wiley & Sons, Ltd.