High Prediction Accuracy Rate Research Articles

Forecasting Core Business Transformation Risk Using the Optimal Rough Set and the Neural Network

AbstractIn the process of enterprise growth, core business transformation is an eternal theme. Enterprise risk forecasting is always an important concern for stakeholders. Considering the completeness and accuracy of the information in the early‐warning index, this paper presents a new risk‐forecasting method for enterprises to use for core business transformation by using rough set theory and an artificial neural network. First, continuous attribute values are discretized using the fuzzy clustering algorithm based on the maximum discernibility value function and information entropy. Afterwards, the major attributes are reduced by the rough sets. The core business transformation risk rank judgement is extracted to define the connection between network nodes and determine the structure of the neural networks. Finally, the improved back‐propagation (BP) neural network learning and training are used to judge the risk level of the test samples. The experiments are based on 265 listed companies in China, and the results show that the proposed risk‐forecasting model based on rough sets and the neural network provides higher prediction accuracy rates than do other widely developed baselines including logistic regression, neural networks and association rules mining. Copyright © 2015 John Wiley & Sons, Ltd.

A QoS-satisfied Prediction Model for Cloud-service Composition Based on Hidden Markov Model

There are various significant issues in cloud computing, such as service provision, service matching and service assessment, which have attracted researchersâ?? attentions recently. QoS play an increasingly important role during the procedure of cloud-based service provision for seamless and dynamic integration of cloud-service components. In this paper, we focus on the QoS-satisfied prediction for cloud-service composited components and present a QoS-satisfied prediction model based on hidden Markov model. For a general process of cloud-service provision, if the userâ??s QoS could not be satisfied only by one cloud-service component, the component composition should be considered to provide to user, where the QoS-satisfied capability of composited components need to be proactively predicted to guarantee the userâ??s QoS. We discuss the proposed model in detail and proof the model partly. The simulation results show that our model can obtain rather high prediction accuracy rate.

DESIGN OF EXPERIMENTS ON NEURAL NETWORK'S PARAMETERS OPTIMIZATION FOR TIME SERIES FORECASTING IN STOCK MARKETS

Artificial neural network (ANN) model has been used for years to conduct research in stock price prediction for three reasons. First, it has a higher prediction accuracy rate in empirical research. Second, it is not subject to the assumption of having samples from a normal distribution. Third, it can deal with non-linear problems. Nevertheless, the accuracy of prediction relies on the parameter settings of neural network as well as the complexities of problems and the neural network architecture; the results of the analysis could be even more significant with the selection of optimal parameters and network architecture. Currently, as a way of setting parameters, most researchers employed the trial and error method. However, this method is very time-consuming and labor-intensive and may not result in the optimal parameters. Therefore, this research took advantage of a back propagation neural network (BPNN) for the purpose of parameter optimization through constructing a model of stock price prediction, applying design of experiment (DOE) to systematize experiment scheduling, and methods of main effects analysis and interaction analysis. The research used two datasets of financial ratios from 50 blue chip companies in Taiwanese stock market and 40 listed American banks in New York stock exchange as experimental samples. Research results showed that the correlation forecasting, root mean squared error (RMSE), and computing time, which can effectively increase the accuracy of stock price prediction, are better than traditional statistical methods and conventional neural network model.

An enhanced hybrid method for time series prediction using linear and neural network models

The need for improving the accuracy of time series prediction has motivated researchers to develop more efficient prediction models. The accuracy rates resulting from linear models such as linear regression (LR), exponential smoothing (ES) and autoregressive integrated moving average (ARIMA) are not high as they are poor in handling the nonlinear time series data. Neural network models are considered to be better in handling such nonlinear time series data. In the real-world problems, the time series data consist of complex linear and nonlinear patterns and it may be difficult to obtain high prediction accuracy rates using only linear or neural network models. Hybrid models which combine both linear and neural network models can be used to obtain high prediction accuracy rates. In this paper, we propose an enhanced hybrid model which indicates for a given input data which choice is better between the two options, namely, a linear-nonlinear combination or a nonlinear-linear combination. The appropriate combination is selected based on a linearity test of data. From the experimental results, it is found that the proposed hybrid model comprising linear-nonlinear combination performs better than other models for the data that have a linear relationship. On the contrary, the hybrid model comprising nonlinear-linear combination performs better than other models for the data that have a nonlinear relationship.

A Meta‐learning Framework for Bankruptcy Prediction

ABSTRACTThe 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.

A Study of Financial Distress based on MDA

The present study is crucial importance to build up a model to develop the predictive abilities for company failures in a later time frame with different financial, business and operating conditions in the Indian context. A total of sixty-four private sector pharmaceutical companies were analyzed with sixteen financial ratios using multiple discriminant analysis. A strong discriminant function was constructed with seven ratios found to be significant in discriminating power and the classification results showed high predictive accuracy rates of between 88% and 94% for each of the five years prior to actual failure. This study also indicated that even with more advanced statistical tools more popularly used recently, MDA is still a very reliable and potent statistical tool.

Knowledge Discovery in Distance Relay Event Report: A Comparative Data-Mining Strategy of Rough Set Theory With Decision Tree

A protective relay performance analysis is only feasible when the hypothesis of expected relay operation characteristics as decision rules is established as the knowledge base. This has been meticulously accomplished by soliciting the relay knowledge domain from protection experts who are usually constrained by their experience and expertise. Manually analyzing an event report is also cumbersome due to the tremendous amount of data to be perused. This paper addresses these issues by intelligently divulging the knowledge hidden in the relay recorded event report using a data-mining strategy based on rough set theory and a rule-quality measure under supervised learning to discover the relay decision algorithm and association rule. The high prediction accuracy rate and the close-to-unity areas under ROC curve value of the relay operating characteristic curve of the discovered relay decision algorithm verifies its generalized ability to predict trip status in an expert system of relay performance analysis. The relay association rule that was subsequently discovered by using the rule-quality analysis had also been verified as being a reliable hypothesis of the relay operation characteristics. This hypothesis helps the protection engineers understand the behavior of the distance relay. These rules would then be compared with and validated by benchmarking decision-tree-based data-mining analysis.

Extração de regras de classificação a partir de redes neurais para auxílio à tomada de decisão na concessão de crédito bancário

A avaliação de risco de crédito é um importante problema administrativo da área de análise financeira. As Redes Neurais têm recebido muita atenção pela sua alta taxa de acurácia preditiva, no entanto não é fácil compreender como elas alcançam as suas decisões. Neste artigo um conjunto de dados de crédito é analisado usando a técnica de extração de regras NeuroRule e o software WEKA para a extração de regras a partir de uma Rede Neural treinada. Os resultados foram considerados bastante satisfatórios alcançando mais de 80% de acurácia quanto à concessão (ou não) de crédito bancário em todas as simulações.

Using Neural Network Rule Extraction and Decision Tables for Credit-Risk Evaluation

Credit-risk evaluation is a very challenging and important management science problem in the domain of financial analysis. Many classification methods have been suggested in the literature to tackle this problem. Neural networks, especially, have received a lot of attention because of their universal approximation property. However, a major drawback associated with the use of neural networks for decision making is their lack of explanation capability. While they can achieve a high predictive accuracy rate, the reasoning behind how they reach their decisions is not readily available. In this paper, we present the results from analysing three real-life credit-risk data sets using neural network rule extraction techniques. Clarifying the neural network decisions by explanatory rules that capture the learned knowledge embedded in the networks can help the credit-risk manager in explaining why a particular applicant is classified as either bad or good. Furthermore, we also discuss how these rules can be visualized as a decision table in a compact and intuitive graphical format that facilitates easy consultation. It is concluded that neural network rule extraction and decision tables are powerful management tools that allow us to build advanced and userfriendly decision-support systems for credit-risk evaluation.

A connectionist approach to generating oblique decision trees

Neural networks and decision tree methods are two common approaches to pattern classification. While neural networks can achieve high predictive accuracy rates, the decision boundaries they form are highly nonlinear and generally difficult to comprehend. Decision trees, on the other hand, can be readily translated into a set of rules. In this paper, we present a novel algorithm for generating oblique decision trees that capitalizes on the strength of both approaches. Oblique decision trees classify the patterns by testing on linear combinations of the input attributes. As a result, an oblique decision tree is usually much smaller than the univariate tree generated for the same domain. Our algorithm consists of two components: connectionist and symbolic. A three-layer feedforward neural network is constructed and pruned, a decision tree is then built from the hidden unit activation values of the pruned network. An oblique decision tree is obtained by expressing the activation values using the original input attributes. We test our algorithm on a wide range of problems. The oblique decision trees generated by the algorithm preserve the high accuracy of the neural networks, while keeping the explicitness of decision trees. Moreover, they outperform univariate decision trees generated by the symbolic approach and oblique decision trees built by other approaches in accuracy and tree size.

Extracting rules from neural networks by pruning and hidden-unit splitting.

An algorithm for extracting rules from a standard three-layer feedforward neural network is proposed. The trained network is first pruned not only to remove redundant connections in the network but, more important, to detect the relevant inputs. The algorithm generates rules from the pruned network by considering only a small number of activation values at the hidden units. If the number of inputs connected to a hidden unit is sufficiently small, then rules that describe how each of its activation values is obtained can be readily generated. Otherwise the hidden unit will be split and treated as output units, with each output unit corresponding to an activation value. A hidden layer is inserted and a new subnetwork is formed, trained, and pruned. This process is repeated until every hidden unit in the network has a relatively small number of input units connected to it. Examples on how the proposed algorithm works are shown using real-world data arising from molecular biology and signal processing. Our results show that for these complex problems, the algorithm can extract reasonably compact rule sets that have high predictive accuracy rates.