Rule-based Decision Tree Research Articles

Integrating Multi-Temporal Spectral and Structural Information from ALOS/AVNIR-2 Images to Map Heterogeneous Land Use/Cover: A Hybrid Approach

Applicability of remotely sensed data in heterogeneous land use/cover mapping has been greatly be restricted, given the fragmented distribution and spectral confusion of land cover features. In addition, ALOS/AVNIR-2 data, which show great potential for land use/cover mapping considering the high-resolution but low-cost characteristics, have not received too much attention. A new hybrid methodology to address these issues is proposed. This approach combines traditional supervised classifiers and unsupervised classifiers and integrates multi-temporal spectral and structural information from ALOS/AVNIR-2 images. The multi-temporal spectral and structural information are then used as auxiliary data through a rule-based decision tree approach to generate a final product with enhanced land use classes and accuracy. A comprehensive evaluation of derived products of the northern part of Shangyu City in eastern coastal China is presented based on official land use/cover map (1:10000) as well as inter-classification consistency analyses. Overall accuracy of 86.5% and Kappa statistics of 0.84 have been achieved, which are significantly higher than those obtained from the Maximum Likelihood classifier and ISODATA classifier. The hybrid approach presented here is straightforward and flexible enough to be generalized so the approach can be applied to interpret similar fragmented land use/cover using various remotely sensed source data.

Household travel attributes transferability analysis: application of a hierarchical rule based approach

This paper presents a system of hierarchical rule-based models of trip generation and modal split. Travel attributes, like trip counts for different transportation modes and commute distance, are among the modeled variables. The proposed framework could be considered as an alternative for several modules of the traditional travel demand modeling approach, while providing travel attributes at the highly disaggregate level that can be also used in activity-based micro-simulation modeling systems. Nonetheless, the modeling framework of this study is not considered as a substitute for activity-based models. The explanatory variables set ranges from socio-economic and demographic attributes of the household to the built environment characteristics of the household residential location. Another important contribution of the study is a framework in which travel attributes are modeled in conjunction with each other and the interdependencies among them are postulated through a hierarchical system of models. All the models are developed using rule-based decision tree method. Moreover, the models developed in this study present a useful improvement in increasing the practicality and accuracy of the rule-based travel data simulation models.

A real-time classification method of power quality disturbances

This paper describes a real-time classification method of power quality (PQ) disturbances. With an acceptable computation burden, both the elementary parameters of the power signal and the types of the disturbances in the power signal are obtained easily. The proposed method addresses the selection of discriminative features for detection and classification of PQ disturbances. Five distinguished time-frequency statistical features of PQ disturbances are extracted using RMS (root-mean-square) method and discrete Fourier transform (DFT). Using a rule-based decision tree (RBDT), the nine types of PQ disturbances can be recognized easily and there is no need to use other complicated classifiers. Finally, the proposed method is tested using the simulated waveforms. And some preliminary experimental results of the accuracy characterization of an initial development instrument are reported. The simulation and application results validate the accuracy and efficiency of the proposed method.

Audio steganalysis with Hausdorff distance higher order statistics using a rule based decision tree paradigm

The aim of this paper is to construct a practical forensic steganalysis tool for audio signals that can properly analyze the statistics disturbed by stego embedding and classify them to selected current steganographic methods. The objective of this paper is to prove that the choice of effective stego sensitive features and a proficient machine learning paradigm enhances the detection accuracy of the steganalyser. In this paper a rule based approach with a family of six decision tree classifiers viz., Alternating Decision Tree, Decision Stump, J48, Logical Model Tree, Naïve Baye’s Tree and Fast Decision Tree learner, to perform the detection of audio subliminal channel is introduced. In particular the higher order statistics extracted from the Hausdorff distance are investigated for an improvement of the detection performance, as competent audio steganalytic features. The evaluation of the enhanced feature space and the decision tree paradigm, on a database containing 4800 clean and stego audio files is performed for classical steganographic as well as for watermarking algorithms. With this strategy it is shown how general forensic approach can detect information hiding techniques in the field of covert communication as well as for DRM applications. For the latter case, the detection of the presence of a potential watermark in a specific feature space can lead to new attacks or to a better design of the watermarking pattern.

Use of a Fast Information Extraction Method as a Decision Support Tool

INTRODUCTION Most electronic documents are available in the free-text format. Analyzability of the free-text information can play a crucial role in decision making through pattern discovery (Lo & Hsieh, 2003). The enormous volume of available documents can be processed semi-automatically by a closed or an open system (Banko & Etzioni, 2008). Important applications such as electronic governance (Zhang, Lin, Lin, & Hsieh, 2008) may entail the use of extracted information from these documents to gain competitiveness and public trust by enforcing transparency. An effectively designed Information Extraction (IE) system can be useful in this regard. An IE system is designed to extract the user-specified items or pre-defined events (Srihari, Li, Niu, & Cornell, 2008) from the text documents of a specific domain. The items to be extracted are specific word(s) of sentences of the input documents. An IE system can be used to fill the fields of a table of a relational database from text documents. Examples of the fields may include the name of a company or type of business. In order to interpret the extracted information, these items can be saved into a relational database. The extracted items can also be used to fill-out forms specified by a user. The information saved in the database can be further processed to identify the relevant correlations. Compared to the extraction of structured data performed by, for example, an ERP system (Wu, Hsieh, Shin, & Wu, 2005), extracting information from a free-text domain is a challenging problem. The reason is that only the contextually relevant target words or phrases have to be extracted. Because statistics does not consider the contexts of the words, the methods such as Naive Bayes Classifier or Average Mutual Information (Carven et al., 2000) are not effective in extracting information. Most of the state of the art IE systems use a combination of statistical and machine learning methods (Freitag, 2000). Jose Iria and Fabio Ciravegna (2006) developed an ontology learning (Suchanek, Sozio, & Weikum, 2009) and document classification method that represents language resources, such as syntactic and semantic parsers, in a way that is independent of the extraction process. Yildiz and Miksch (2007) proposed an unsupervised rule learning method (Downey, Schoenmackers, & Etzioni, 2007; Sekine & Oda, 2007) based on ontological structures. Syntactic and semantic analyses have been found to improve precision (Feldman, Aumann, Finkelstein-Landau, Hurvitz, Regev, & Yaroshevich, 2002). Xu, Uszkoreit, Li, and Felger (2008) designed a system that extracts linguistic grammar rules from a semantic seed. An alignment based pattern matching technique was developed by (Kim, Jeong, Lee, Ko, & Lee, 2008) that can extract relationships between two arguments. Jain, Ipeirotis, and Gravano (2008) developed a technique to process structured queries from the relations extracted from unstructured texts. Stevenson and Greenwood (2009) found that the IE models that use the relevant portions of a dependency tree (Wu & Weld, 2008) perform better. A rule-based decision tree was found effective in extracting information from online resumes (Bhargavi, Jyothi, Jyothi, & Sekar, 2008). The use of deep level ontology structure (Welty & Murdock, 2006) and the use of multiple ontologies from the same domain (Wimalasuriya & Dou, 2009) have been found effective in performance enhancement. Assigning weights to the syntactic features according to their co-occurrences in the related class has been found effective in determining ontologies about person-names and their geographic locations (Tanev & Magnini, 2008). Enterprise applications of IE systems require higher accuracy and scalability (Chiticariu, Li, Raghavan, & Reiss, 2010). Similar to many other systems, Pennacchiotti and Pantel (2009) used a combination of a pattern match method with a distributional method. …

Spatial decision support system for Medfly control in citrus

A spatial decision support system (SDSS), designated MedCila was developed for controlling Medfly on citrus in Israel. The development involved four main phases: (1) acquisition of relevant expert and domain knowledge; (2) identification of the relevant criteria and modeling each criterion and the overall decision-making procedure; (3) integration of the MedCila into a GIS environment; and (4) initial evaluation of MedCila performance. The criteria found to be most relevant for control decision-making were the number of flies and the presence of a ‘blue eye’ in the nearest trap, the host-species susceptibility, the relative development of the Medfly based on accumulative day-degree model, the history of trapping, and the Medfly population in the nearby traps. Binary, linear, logarithmic and biological-based models were developed for the criteria identified. The overall decision-making procedure of the MedCila was based on the Stanford Certainty Theory integrated with a rule-based decision tree. Initial evaluation of the MedCila performance was done by retrospective comparison between the MedCila recommendations and the coordinator decisions. It was shown that the MedCila provides recommendations that are generally accepted by the coordinators; it reduces the number of unnecessary spray actions in the absence of a Medfly threat in space and in time; and it reduces the number of plots for which the coordinator needs to make a decision for.

Vegetation water content during SMEX04 from ground data and Landsat 5 Thematic Mapper imagery

Vegetation water content is an important parameter for retrieval of soil moisture from microwave data and for other remote sensing applications. Because liquid water absorbs in the shortwave infrared, the normalized difference infrared index (NDII), calculated from Landsat 5 Thematic Mapper band 4 (0.76–0.90 μm wavelength) and band 5 (1.55–1.65 μm wavelength), can be used to determine canopy equivalent water thickness (EWT), which is defined as the water volume per leaf area times the leaf area index (LAI). Alternatively, average canopy EWT can be determined using a landcover classification, because different vegetation types have different average LAI at the peak of the growing season. The primary contribution of this study for the Soil Moisture Experiment 2004 was to sample vegetation for the Arizona and Sonora study areas. Vegetation was sampled to achieve a range of canopy EWT; LAI was measured using a plant canopy analyzer and digital hemispherical (fisheye) photographs. NDII was linearly related to measured canopy EWT with an R 2 of 0.601. Landcover of the Arizona, USA, and Sonora, Mexico, study areas were classified with an overall accuracy of 70% using a rule-based decision tree using three dates of Landsat 5 Thematic Mapper imagery and digital elevation data. There was a large range of NDII per landcover class at the peak of the growing season, indicating that canopy EWT should be estimated directly using NDII or other shortwave-infrared vegetation indices. However, landcover classifications will still be necessary to obtain total vegetation water content from canopy EWT and other data, because considerable liquid water is contained in the non-foliar components of vegetation.

The cost of implementing UK guidelines for the management of chronic kidney disease

Chronic kidney disease (CKD) is a major public health problem. In the UK, guidelines have been developed to facilitate case identification and management. Our aim was to estimate the annualized cost of implementation of the guidelines on newly identified CKD cases. We interrogated the New Opportunities for Early Renal Intervention by Computerised Assessment (NEOERICA) database using a Java program created to recompile the CKD guidelines into rule-based decision trees. This categorized all patients with a serum creatinine recorded over a 1-year period into those requiring more tests or referral. A 12-month cost analysis for following the guidelines was performed. In the first year, a practice of 10,000 would identify 147.5 patients with stages 3-5 CKD over and above those already known. All stages 4-5 CKD cases would require nephrology referral. Of those with stage 3 CKD (143.85), 126.27 stable patients would require more tests. The following would require referral: 14.8 with estimated glomerular filtration rate decline>or=5 ml/min/1.73 m2/year, 1.11 with haemoglobin<11 g/dl and 1.67 with blood pressure>150/90 on three anti-hypertensives. The projected cost per practice of investigating stable stage 3 CKD was euro 6111; and euro 7836 for nephrology referral. Total costs of euro 17 133 in the first year were increased to euro 29,790 through the effect of creatinine calibration. CKD guideline implementation results in significant increases in nephrology referral and additional investigation. These costs could be recouped by delaying dialysis requirement by 1 year in one individual per 10,000 patients managed according to guidelines.

Power-Quality Disturbance Recognition Using S-Transform

Taking advantage of S-transforms (STs), this paper proposes a new method of detecting and classifying power-quality disturbances. The ST is unique in that it provides frequency-dependent resolution while maintaining a direct relationship with the Fourier spectrum. The features obtained from ST are distinct, understandable, and immune to noise. According to a rule-based decision tree, eight types of single power disturbance and two types of complex power disturbance are well recognized, and there is no need to use other complicated classifiers. The comparison between the wavelet-transform-based method and the ST-based method for power-quality disturbance recognition is also provided. The simulation results show that the proposed method is effective and immune against noise. The proposed method is feasible and promising for real applications

Detection of e-mail concerning criminal activities using association rule-based decision tree

Detection of e-mails about criminal activities using association rule-based decision tree is studied here. Instead of using words, word-relation, that is, association rules from these words, is used for building decision tree. In our experiments, we first preprocess data. We then find out association relations among these words using Rakesh Agrawal et al.'s Apriori algorithm applying objective interestingness measures. These rules are used for training and testing the decision tree-based classification system. A discussion of the result obtained is also given.

Risk Factors to Retrieve Anomaly Intrusion Information and Profile User Behavior

The use of network traffic audit data for retrieving anomaly intrusion information and profiling user behavior has been studied previously, but the risk factors associated with attacks remain unclear. This study aimed to identify a set of robust risk factors via the bootstrap resampling and logistic regression modeling methods based on the KDD-cup 1999 data. Of the 46 examined variables, 16 were identified as robust risk factors, and the classification showed similar performances in sensitivity, specificity, and correctly classified rate in comparison with the KDD-cup 1999 winning results that were based on a rule-based decision tree algorithm with all variables. The study emphasizes that the bootstrap simulation and logistic regression modeling techniques offer a novel approach to understanding and identifying risk factors for better information protection on network security.

A multinomial logistic regression modeling approach for anomaly intrusion detection

Although researchers have long studied using statistical modeling techniques to detect anomaly intrusion and profile user behavior, the feasibility of applying multinomial logistic regression modeling to predict multi-attack types has not been addressed, and the risk factors associated with individual major attacks remain unclear. To address the gaps, this study used the KDD-cup 1999 data and bootstrap simulation method to fit 3000 multinomial logistic regression models with the most frequent attack types ( probe, DoS, U2R, and R2L) as an unordered independent variable, and identified 13 risk factors that are statistically significantly associated with these attacks. These risk factors were then used to construct a final multinomial model that had an ROC area of 0.99 for detecting abnormal events. Compared with the top KDD-cup 1999 winning results that were based on a rule-based decision tree algorithm, the multinomial logistic model-based classification results had similar sensitivity values in detecting normal (98.3% vs. 99.5%), probe (85.6% vs. 83.3%), and DoS (97.2% vs. 97.1%); remarkably high sensitivity in U2R (25.9% vs. 13.2%) and R2L (11.2% vs. 8.4%); and a significantly lower overall misclassification rate (18.9% vs. 35.7%). The study emphasizes that the multinomial logistic regression modeling technique with the 13 risk factors provides a robust approach to detect anomaly intrusion.

Classification of coral reef images from underwater video using neural networks

We use a feedforward backpropagation neural network to classify close-up images of coral reef components into three benthic categories: living coral, dead coral and sand. We have achieved a success rate of 86.5% (false positive = 6.7%) for test images that were not in the training set which is high considering that corals occur in an immense variety of appearance. Color and texture features derived from video stills of coral reef transects from the Great Barrier Reef were used as inputs to the network. We also developed a rule-based decision tree classifier according to how marine scientists classify corals from texture and color, and obtained a lower recognition rate of 79.7% for the same set of images.

Added value of new acute coronary syndrome computer algorithm for interpretation of prehospital electrocardiograms

A new computerized acute coronary syndrome (ACS) computer algorithm has been developed with the aim of improving the electrocardiographic detection of acute myocardial ischemia and infarction in the emergency department (ED). The purpose of this study was to determine the added value of the new ACS algorithm in assisting ED physicians to obtain a more accurate diagnosis in patients with ACS. The new algorithm combines a rule-based decision tree, which uses well-known clinical criteria and a data-centered neural network model for more robust pattern recognition. Input parameters of the neural network model consist of morphology features of derived Frank X, Y, Z waveforms and the patient’s gender and age. The neural network model was trained with electrocardiograms obtained from documented acute myocardial infarction patients at the Mayo Clinic who were a part of a research ACS database, which includes electrocardiograms (ECGs) of more than 5,000 individuals at hospital admission (1 st ECG in the ED). The test set portion of the study was conducted in 2 steps: 1) One emergency physician and 1 cardiologist classified 1,902 clinically correlated out-of-hospital ECGs without seeing the interpretation statement from the algorithm into 1 of the following categories: 1) acute myocardial infarction, acute ischemia, or nonischemic; 2) After 9 months, the same 2 physicians classified the same group of ECGs but with the interpretation statement of the algorithm printed on the tracing. The results demonstrated that with the assistance of the new algorithm, the emergency physician and cardiologist improved their sensitivity of interpreting acute myocardial infarction by 50% and 26%, respectively, without a loss of specificity. The new algorithm also improved the emergency physician’s acute ischemia interpretation sensitivity by 53% and still maintained a reasonable specificity (91%). The new ACS algorithm provides added value for improving acute ischemia and infarction detection in the ED.

Policy Relevant Modelling: Relationships Between Water, Land Use, and Farmer Decision Processes

This paper presents a policy-relevant model based approach to assessing land-use change that is based upon a number of transdisciplinary mechanisms. We report the development of a modelling platform which supports analysis of the interactions between slope, aquifer, and catchment dynamics, together with the actions of farmers as they change their crop profiles and consequently their water needs. The master-equation based farmer decision making model, driven by rule based decision trees derived within the context of a conceptual framework which accommodates both the human and natural processes and their interdependencies, is central to the method presented in this paper. Our conclusions relate to both the function and process of integrative assessment and we present characteristic model outputs to highlight the interactions between the farmers and their natural environment.

The role of expert system shell induction in the analysis of phase and amplitude images obtained from nuclear cardiology

In an attempt to explain the results of a perception experiment, estimates of statistical parameters were obtained from phase and amplitude images of nuclear medicine multiple-gated cardiac studies. A commercially available induction system was used to create a rule-based decision tree using these estimates as a training set. In the classification of normal and abnormal studies two useful rules were found; one of these involved the standard deviation of the left ventricular phase histogram and the other concerned the correlation of adjacent pixels in the left ventricular region of the amplitude image. A discriminant analysis performed on five amplitude and five phase parameters produced no better discrimination than the use of the phase standard deviation alone. Although it was hoped that induction would find more than two useful rules, the results of the discriminant analysis confirm that induction has performed creditably on a difficult problem.