A Meta‐learning Framework for Bankruptcy Prediction

The implication of corporate bankruptcy prediction is important to financial institutions when making lending decisions. In related studies, many bankruptcy prediction models have been developed based on some machine‐learning techniques. This paper presents a meta‐learning framework, which is composed of two‐level classifiers for bankruptcy prediction. The first‐level multiple classifiers perform the data reduction task by filtering out unrepresentative training data. Then, the outputs of the first‐level classifiers are utilized to create the second‐level single (meta) classifier. The experiments are based on five related datasets and the results show that the proposed meta‐learning framework provides higher prediction accuracy rates and lower type I/II errors when compared with the stacked generalization classifier and other three widely developed baselines, such as neural networks, decision trees, and logistic regression. Copyright © 2011 John Wiley & Sons, Ltd.     

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.     

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.     

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.     

Analysis of the relationship between LSTM network traffic flow prediction performance and statistical characteristics of standard and nonstandard data

The effectiveness of road traffic control systems can be increased with the help of a model that can accurately predict short-term traffic flow. Therefore, the performance of the preferred approach to develop a prediction model should be evaluated with data sets with different statistical characteristics. Thus a correlation can be established between the statistical properties of the data set and the model performance. The determination of this relationship will assist experts in choosing the appropriate approach to develop a high-performance short-term traffic flow forecasting model. The main purpose of this study is to reveal the relationship between the long short-term memory network (LSTM) approach's short-term traffic flow prediction performance and the statistical properties of the data set used to develop the LSTM model. In order to reveal these relationships, two different traffic prediction models with LSTM and nonlinear autoregressive (NAR) approaches were created using different data sets, and statistical analyses were performed. In addition, these analyses were repeated for nonstandardized traffic data indicating unusual fluctuations in traffic flow. As a result of the analyses, LSTM and NAR model performances were found to be highly correlated with the kurtosis and skewness changes of the data sets used to train and test these models. On the other hand, it was found that the difference of mean and skewness values of training and test sets had a significant effect on model performance in the prediction of nonstandard traffic flow samples.     

Forecasting Monetary Policy Decisions in Australia: A Forecast Combinations Approach

This paper applies a triple‐choice ordered probit model, corrected for nonstationarity to forecast monetary decisions of the Reserve Bank of Australia. The forecast models incorporate a mix of monthly and quarterly macroeconomic time series. Forecast combination is used as an alternative to one multivariate model to improve accuracy of out‐of‐sample forecasts. This accuracy is evaluated with scoring functions, which are also used to construct adaptive weights for combining probability forecasts. This paper finds that combined forecasts outperform multivariable models. These results are robust to different sample sizes and estimation windows. Copyright © 2011 John Wiley & Sons, Ltd.     

Assessment of agricultural energy consumption of Turkey by MLR and Bayesian optimized SVR and GPR models

Agricultural productivity highly depends on the cost of energy required for cultivation. Thus prior knowledge of energy consumption is an important step for energy planning and policy development in agriculture. The aim of the present study is to evaluate the application potential of multiple linear regression (MLR) and machine learning tools such as support vector regression (SVR) and Gaussian process regression (GPR) to forecast the agricultural energy consumption of Turkey. In the development of the models, widespread indicators such as agricultural value-added, total arable land, gross domestic product share of agriculture, and population data were used as input parameters. Twenty-eight-year historical data from 1990 to 2017 were utilized for the training and testing stages of the models. A Bayesian optimization method was applied to improve the prediction capability of SVR and GPR models. The performance of the models was measured by various statistical tools. The results indicated that the Bayesian optimized GPR (BGPR) model with exponential kernel function showed a superior prediction capability over MLR and Bayesian optimized SVR model. The root mean square error, mean absolute deviation, mean absolute percentage error, and coefficient of determination (R²) values for the BGPR model were determined as 0.0022, 0.0005, 0.2041, and 0.9999 in the training phase and 0.0452, 0.0310, 7.7152, and 0.9677 in the testing phase, respectively. As a result, it can be concluded that the proposed BGPR model is an efficient technique and has the potential to predict agricultural energy consumption with high accuracy.     

Detecting and Predicting Economic Accelerations,Recessions, and Normal Growth Periods in Real‐Time

The dichotomous characterization of the business cycle in recessions and expansions has been central in the literature over the last 50 years. However, there are various reasons to question the adequacy of this dichotomous, recession/expansion approach for our understanding of the business cycle dynamics, as well as for the prediction of future business cycle developments. In this context, the contribution of this paper to the literature is twofold. First, since a positive rate of growth at the level of economic activity can be considered as the normal scenario in modern economies due to both population and technological growth, it proposes a new non‐parametric algorithm for the detection and dating of economic acceleration periods, trend or normal growth periods, and economic recessions. Second, it uses an ordered probit framework for the estimation and forecasting of these three business cycle phases, applying an automatized model selection approach using monthly macroeconomic and financial data on the German economy. The empirical results show that this approach has superior out‐of‐sample properties under real‐time conditions compared to alternative probit models specified individually for the prediction of recessions and/or economic accelerations. Copyright © 2016 John Wiley & Sons, Ltd.     

Forecasting football results and the efficiency of fixed‐odds betting

An ordered probit regression model estimated using 10 years' data is used to forecast English league football match results. As well as past match results data, the significance of the match for end‐of‐season league outcomes, the involvement of the teams in cup competition and the geographical distance between the two teams' home towns all contribute to the forecasting model's performance. The model is used to test the weak‐form efficiency of prices in the fixed‐odds betting market. A strategy of selecting end‐of‐season bets with a favourable expected return according to the model appears capable of generating a positive return. Copyright © 2004 John Wiley & Sons, Ltd.     

Comparison of two non‐parametric models for daily traffic forecasting in Hong Kong

The most up‐to‐date annual average daily traffic (AADT) is always required for transport model development and calibration. However, the current‐year AADT data are not always available. The short‐term traffic flow forecasting models can be used to predict the traffic flows for the current year. In this paper, two non‐parametric models, non‐parametric regression (NPR) and Gaussian maximum likelihood (GML), are chosen for short‐term traffic forecasting based on historical data collected for the annual traffic census (ATC) in Hong Kong. These models are adapted as they are more flexible and efficient in forecasting the daily vehicular flows in the Hong Kong ATC core stations (in total of 87 stations). The daily vehicular flows predicted by these models are then used to calculate the AADT of the current year, 1999. The overall prediction and comparison results show that the NPR model produces better forecasts than the GML model using the ATC data in Hong Kong. Copyright © 2006 John Wiley _ Sons, Ltd.     

Removing Forecasting Errors with White Gaussian Noise after Square Root Transformation

An analytical model has been developed in the present paper based on a square root transformation of white Gaussian noise. The mathematical expectation and variance of the new asymmetric distribution generated by white Gaussian noise after a square root transformation are analytically deduced from the preceding four terms of the Taylor expansion. The model was first evaluated against numerical experiments and a good agreement was obtained. The model was then used to predict time series of wind speeds and highway traffic flows. The simulation results from the new model indicate that the prediction accuracy could be improved by 0.1–1% by removing the mean errors. Further improvement could be obtained for non‐stationary time series, which had large trends. Copyright © 2016 John Wiley & Sons, Ltd.     

Short‐term prediction of motorway travel time using ANPR and loop data

Travel time is a good operational measure of the effectiveness of transportation systems. The ability to accurately predict motorway and arterial travel times is a critical component for many intelligent transportation systems (ITS) applications. Advanced traffic data collection systems using inductive loop detectors and video cameras have been installed, particularly for motorway networks. An inductive loop can provide traffic flow at its location. Video cameras with image‐processing software, e.g. Automatic Number Plate Recognition (ANPR) software, are able to provide travel time of a road section. This research developed a dynamic linear model (DLM) model to forecast short‐term travel time using both loop and ANPR data. The DLM approach was tested on three motorway sections in southern England. Overall, the model produced good prediction results, albeit large prediction errors occurred at congested traffic conditions due to the dynamic nature of traffic. This result indicated advantages of use of the both data sources. Copyright © 2008 John Wiley & Sons, Ltd.     

Forecasting US Recessions with a Large Set of Predictors

In this paper, I use a large set of macroeconomic and financial predictors to forecast US recession periods. I adopt Bayesian methodology with shrinkage in the parameters of the probit model for the binary time series tracking the state of the economy. The in‐sample and out‐of‐sample results show that utilizing a large cross‐section of indicators yields superior US recession forecasts in comparison to a number of parsimonious benchmark models. Moreover, the data‐rich probit model gives similar accuracy to the factor‐based model for the 1‐month‐ahead forecasts, while it provides superior performance for 1‐year‐ahead predictions. Finally, in a pseudo‐real‐time application for the Great Recession, I find that the large probit model with shrinkage is able to pick up the recession signals in a timely fashion and does well in comparison to the more parsimonious specification and to nonparametric alternatives. Copyright © 2016 John Wiley & Sons, Ltd.     

Predicting multistage financial distress: Reflections on sampling,feature and model selection criteria

This research proposes a prediction model of multistage financial distress (MSFD) after considering contextual and methodological issues regarding sampling, feature and model selection criteria. Financial distress is defined as a three‐stage process showing different nature and intensity of financial problems. It is argued that applied definition of distress is independent of legal framework and its predictability would provide more practical solutions. The final sample is selected after industry adjustments and oversampling the data. A wrapper subset data mining approach is applied to extract the most relevant features from financial statement and stock market indicators. An ensemble approach using a combination of DTNB (decision table and naïve base hybrid model), LMT (logistic model tree) and A2DE (alternative N dependence estimator) Bayesian models is used to develop the final prediction model. The performance of all the models is evaluated using a 10‐fold cross‐validation method. Results showed that the proposed model predicted MSFD with 84.06% accuracy. This accuracy increased to 89.57% when a 33.33% cut‐off value was considered. Hence the proposed model is accurate and reliable to identify the true nature and intensity of financial problems regardless of the contextual legal framework.     

人体生物力学模型建立及在汽车安全中的应用

汽车的发明给人类社会带来了巨大进步和便利,但同时引发的交通事故造成大量的人员伤亡与经济损失.本论文采用医学CT/MRI扫描获得人体主要组织的几何参数,采用有限元仿真分析工具建立人体有限元生物力学模型,结合文献已有的组织材料实验与尸体实验对模型进行仿真验证,模型验证参考了尸体实验数据进行了正面碰撞和侧面碰撞的仿真验证,碰撞中接触力、胸骨位移量、力-位移响应与实验吻合较好.开展了汽车方向盘与胸部碰撞的仿真应用研究,方向盘冲击作用于人体的生物力学响应与损伤特性符合真实事故中的变形规律,能再现肋骨骨折及内脏损伤,模型可应用于汽车碰撞事故中人员的损伤评估研究.     

Unemployment variation over the business cycles: a comparison of forecasting models

Asymmetry has been well documented in the business cycle literature. The asymmetric business cycle suggests that major macroeconomic series, such as a country's unemployment rate, are non‐linear and, therefore, the use of linear models to explain their behaviour and forecast their future values may not be appropriate. Many researchers have focused on providing evidence for the non‐linearity in the unemployment series. Only recently have there been some developments in applying non‐linear models to estimate and forecast unemployment rates. A major concern of non‐linear modelling is the model specification problem; it is very hard to test all possible non‐linear specifications, and to select the most appropriate specification for a particular model. Artificial neural network (ANN) models provide a solution to the difficulty of forecasting unemployment over the asymmetric business cycle. ANN models are non‐linear, do not rely upon the classical regression assumptions, are capable of learning the structure of all kinds of patterns in a data set with a specified degree of accuracy, and can then use this structure to forecast future values of the data. In this paper, we apply two ANN models, a back‐propagation model and a generalized regression neural network model to estimate and forecast post‐war aggregate unemployment rates in the USA, Canada, UK, France and Japan. We compare the out‐of‐sample forecast results obtained by the ANN models with those obtained by several linear and non‐linear times series models currently used in the literature. It is shown that the artificial neural network models are able to forecast the unemployment series as well as, and in some cases better than, the other univariate econometrics time series models in our test. Copyright © 2004 John Wiley & Sons, Ltd.     

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.     

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.     

A Bayesian structural model for predicting algal blooms

A Bayesian structural model with two components is proposed to forecast the occurrence of algal blooms, multivariate mean‐reverting diffusion process (MMRD), and a binary probit model with latent Markov regime‐switching process (BPMRS). The model has three features: (a) forecast of the occurrence probability of algal bloom is directly based on oceanographic parameters, not the forecasting of special indicators in traditional approaches, such as phytoplankton or chlorophyll‐a; (b) augmentation of daily oceanographic parameters from the data collected every 2 weeks is based on MMRD. The proposed method solves the problem of unavailability of daily oceanographic parameters in practice; (c) BPMRS captures the unobservable factors which affect algal bloom occurrence and therefore improve forecast accuracy. We use panel data collected in Tolo Harbour, Hong Kong, to validate the model. The model demonstrates good forecasting for out‐of‐sample rolling forecasts, especially for algal bloom appearing for a longer period, which severely damages fisheries and the marine environment.     

Using a heterogeneous multinomial probit model with a neural net extension to model brand choice

The multinomial probit model introduced here combines heterogeneity across households with flexibility of the (deterministic) utility function. To achieve flexibility deterministic utility is approximated by a neural net of the multilayer perceptron type. A Markov Chain Monte Carlo method serves to estimate heterogeneous multinomial probit models which fulfill economic restrictions on signs of (marginal) effects of predictors (e.g., negative for price). For empirical choice data the heterogeneous multinomial probit model extended by a multilayer perceptron clearly outperforms all the other models studied. Moreover, replacing homogeneous by heterogeneous reference price mechanisms and thus allowing price expectations to be formed differently across households also leads to better model performance. Mean utility differences and mean elasticities w.r.t. price and price deviation from reference price demonstrate that models with linear utility and nonlinear utility approximated by a multilayer perceptron lead to very different implications for managerial decision making. Copyright © 2007 John Wiley & Sons, Ltd.