Search results

Explores, using appropriate examples, the ways in which decision trees can be used by the manager to assist in the longitudinal decision‐making process. Since the mathematical concepts associated with decision trees are complex, managers can be reluctant to attempt to use decision tree models. A recognition that such models can be simply developed in a spreadsheet environment, and can then be used for sensitivity analysis using different decision criteria, demonstrates that decision trees can offer valuable insights into the structure of a strategic decision problem.

Email spam classification is now becoming a challenging area in the domain of text classification. Precise and robust classifiers are not only judged by classification accuracy but also by sensitivity (correctly classified legitimate emails) and specificity (correctly classified unsolicited emails) towards the accurate classification, captured by both false positive and false negative rates. This paper aims to present a comparative study between various decision tree classifiers (such as AD tree, decision stump and REP tree) with/without different boosting algorithms (bagging, boosting with re-sample and AdaBoost).

Artificial intelligence and text mining approaches have been incorporated in this study. Each decision tree classifier in this study is tested on informative words/features selected from the two publically available data sets (SpamAssassin and LingSpam) using a greedy step-wise feature search method.

Outcomes of this study show that without boosting, the REP tree provides high performance accuracy with the AD tree ranking as the second-best performer. Decision stump is found to be the under-performing classifier of this study. However, with boosting, the combination of REP tree and AdaBoost compares favourably with other classification models. If the metrics false positive rate and performance accuracy are taken together, AD tree and REP tree with AdaBoost were both found to carry out an effective classification task. Greedy stepwise has proven its worth in this study by selecting a subset of valuable features to identify the correct class of emails.

This research is focussed on the classification of those email spams that are written in the English language only. The proposed models work with content (words/features) of email data that is mostly found in the body of the mail. Image spam has not been included in this study. Other messages such as short message service or multi-media messaging service were not included in this study.

In this research, a boosted decision tree approach has been proposed and used to classify email spam and ham files; this is found to be a highly effective approach in comparison with other state-of-the-art modes used in other studies. This classifier may be tested for different applications and may provide new insights for developers and researchers.

A comparison of decision tree classifiers with/without ensemble has been presented for spam classification.

For classification problems of customer relationship management (CRM), the purpose of this paper is to propose a method with interpretability of the classification results that combines multiple decision trees based on a genetic algorithm.

In the proposed method, multiple decision trees are combined in parallel. Subsequently, a genetic algorithm is used to optimize the weight matrix in the combination algorithm.

The method is applied to customer credit rating assessment and customer response behavior pattern recognition. The results demonstrate that compared to a single decision tree, the proposed combination method improves the predictive accuracy and optimizes the classification rules, while maintaining interpretability of the classification results.

The findings of this study contribute to research methodologies in CRM. It specifically focuses on a new method with interpretability by combining multiple decision trees based on genetic algorithms for customer classification.

Condition‐based maintenance (CBM) has increasingly drawn attention in industry because of its many benefits. The CBM problem is a kind of state‐dependent scheduling problem, and is very hard to solve within the conventional Markov decision process framework. The purpose of this paper is to present an intelligent CBM scheduling model for which incremental decision tree learning as an evolutionary system identification model and dynamic programming as a control model are developed.

To fully exploit the merits of CBM, this paper models CBM scheduling as a state‐dependent, sequential decision‐making problem. The objective function is formulated as the minimization of the total maintenance cost. Instead of interpreting the problem within the widely used Markovian framework, this paper proposes an intelligent maintenance scheduling approach that integrates an incremental decision tree learning method and deterministic dynamic programming techniques.

Although the intelligent maintenance scheduling approach proposed in this paper does not guarantee an optimal scheduling policy from a mathematical viewpoint, it is verified through a simulation‐based experiment that the intelligent maintenance scheduler is capable of providing a good scheduling policy that can be used in practice.

This paper presents an intelligent maintenance scheduler. As a system identification model, we devise a new incremental decision tree learning method by which interaction patterns among attributes and machine condition are disclosed in an evolutionary manner. A deterministic dynamic programming technique is then applied to select the best safe state in terms of the total maintenance cost.

This paper aims to assist investigators and attorneys addressing the legal aspects of cyber incidents, and allow them to determine the legality of a response to cyber attacks by using the Worldwide web securely.

Develop a decision support legal whiteboard that graphically constructs legal arguments as a decision tree. The tree is constructed using a tree of questions and appending legal documents to substantiate the answers that are known to hold in anticipated legal challenges.

The tool allows participating group of attorneys to meet in cyberspace in real time and construct a legal argument graphically by using a decision tree. They can construct sub‐parts of the tree from their own legal domains. Because diverse legal domains use different nomenclatures, this tool provides the user the capability to index and search legal documents using a complex international legal ontology that goes beyond the traditional LexisNexis‐like legal databases. This ontology itself can be created using the tool from distributed locations.

This tool has been fine‐tuned through numerous interviews with attorneys teaching and practicing in the area of cyber crime, cyber espionage, and military operations in cyberspace. It can be used to guide forensic experts and law enforcement personnel during their active responses and off‐line examinations.

The purpose of this paper is to investigate the use of decision tree analysis in the identification of stakeholders based on their participation in tourism and political activities in a community.

A survey was sent to tourism stakeholders in two rural counties. Responses were collected and analyzed using the exhaustive chi‐square automatic interaction detection decision tree analysis.

Based on the results of the decision tree analysis four tourism stakeholder groups were identified based on their participation in tourism and political activities in a community: high participants, high‐moderate participants, low‐moderate participants, and low participants.

Owing to a low response rate, an issue of non‐response bias could exist, but the information from the respondents can give insight on stakeholders in these communities. Also, the specific results of this study can only be applied to eastern North Carolina and are not generalizable to other areas.

Results from this study demonstrate the use of decision tree analysis in identifying community stakeholders. Using decision tree analysis tourism planners can identify stakeholder groups that will participate in tourism and political activities. With this knowledge, tourism planners can identify which stakeholder groups will be the most influential and vocal in a community with regard to tourism development.

Decision tree analysis is a tool for partitioning a data set based on the relationships between a set of independent variables and a dependent variable. The research reported here tests the application of decision tree analysis, an analytical technique that is not traditionally used to segment stakeholders in tourism.

The purpose of this paper is to provide classification confidence value to every individual sample classified by decision trees and use this value to combine the classifiers.

The proposed system is first theoretically explained, and then the use and effectiveness of the proposed system is demonstrated on sample datasets.

In this paper, a novel method is proposed to combine decision tree classifiers using calculated classification confidence values. This confidence in the classification is based on distance calculation to the relevant decision boundary (distance conditional), probability density estimation and (distance conditional) classification confidence estimation. It is shown that these values – provided by individual classification trees – can be integrated to derive a consensus decision.

The proposed method is not limited to axis‐parallel trees, it is applicable not only to oblique trees, but also to any kind of classifier system that uses hyperplanes to cluster the input space.

A novel method is presented to extend decision tree like classifiers with confidence calculation and a voting system is proposed that uses this confidence information. The proposed system possesses several novelties (e.g. it not only gives class probabilities, but also classification confidences) and advantages over previous (traditional) approaches. The voting system does not require an auxiliary combiner or gating network, as in the mixture of experts structure and the method is not limited to decision trees with axis‐parallel splits; it is applicable to any kind of classifiers that use hyperplanes to cluster the input space.

The purpose of this paper is to explore teachers' behaviours in completing an internet survey using decision trees. Furthermore, to reduce the complexity of the decision trees, a statistical technique was used to decrease the number of input variables in the decision trees.

A dataset of 47,647 samples was used to build the decision trees. These samples were collected from an internet survey of teachers in Taiwan. The output of the decision trees was the answering time (the time taken to complete the internet questionnaire). Eight variables were selected as the inputs for the decision trees. Two techniques were employed to build the decision trees – the exhaustive chi‐squared automatic interaction detector (ECHAID) and classification and regression tree (CRT) analysis. To reduce the complexity of the decision models, factor analysis technique was used to decrease the data dimensions (number of input variables) and to obtain a simplified decision model. One‐way ANOVA was used to validate the effects of the dimension reduction.

From the results of the factor analysis, a simplified decision tree is recommended using four input variables – teaching years, school level, sex and area. The classification accuracy of the simplified model is statistically equivalent to that of the original one, which used eight input variables.

The complexity of decision trees theoretically depends on the number of input variables. This study used a statistical technique to decrease the number of input variables and thereby reduce the complexity of the decision trees. A statistical technique was employed to validate that the classification accuracy is not statistically different between the original decision model and the simplified one. The decision models proposed in this paper can be applied in estimating the answering time for completing a questionnaire during an internet survey.

The purpose of this paper is to analyze medium‐term risks faced by electrical generation companies in competitive environments. Market risks faced by generation companies are caused by several variables subject to uncertainty. Hydro conditions, fuel (coal and natural gas) prices, system demand, and CO₂ emission price are the risk factors considered in the paper. Taking into account these risk factors, generation companies have to take decisions that would affect their economic results and their risk exposure.

This paper proposes a methodology to support the risk‐analysis decision‐making process. Firstly, different scenarios of risk factors are generated. Then, a market equilibrium model is used in order to assess the impact of the different sources of uncertainty. Finally, decision trees are used in order to analyze the variables subject to interest, such as electricity prices or companies' profits.

The proposed methodology can be enhanced to take into account scenarios of more risk factors, such as equipment failure or agents' behavior. Another future enhancement could be a detailed study of correlation between different risk factors.

A realistic case study is presented, showing the advantages of these techniques for medium‐term risk‐analysis and decision‐making processes. Several decision trees have been generated to assess the impact of the different risk factors in electricity prices and companies' profits. These decision trees provide valuable information for companies when facing their risk‐management process.

The approach presented here constitutes a valuable support to gain useful information for wise decision making and to hedge against risk.

Determining the relevant information using financial measures is of great interest for various stakeholders to analyze the performance of the firm. This paper aims at identifying these financial measures (ratios) which critically affect the firm performance. The authors specifically focus on discovering the most prominent ratios using a two-step process. First, the authors use an exploratory factor analysis to identify the underlying dimensions of these ratios, followed by predictive modeling techniques to identify the potential relationship between measures and performance.

The study uses data of 25 financial variables for a sample of 1923 Indian manufacturing firms which exist continuously between 2011 and 2018. For prediction models, four popular decision tree algorithms [Chi-squared automatic interaction detector (CHAID), classification and regression trees (C&RT), C5.0 and quick, unbiased, efficient statistical tree (QUEST)] were investigated, and the information fusion-based sensitivity analyses were performed to identify the relative importance of these input measures.

Results show that C5.0 and CHAID algorithms produced the best predictive results. The fusion sensitivity results find that net profit margin and total assets turnover rate are the most critical factors determining the firm performance in an Indian manufacturing context. These findings may enable managers in their decision-making process and also have vital implications for investors in assessing the performance of the firm.

To the best of the authors’ knowledge, the current paper is the first to address the application of decision tree algorithms to predict the performance of manufacturing firms in an emerging economy such as India, with the latest data. This practical perspective helps the organizations in managing the critical parameters for the firm’s growth.