Chien-Chung Chan

A rough set approach to attribute generalization in data mining

Abstract This paper presents a method for updating approximations of a concept incrementally. The results can be used to implement a quasi-incremental algorithm for learning classification rules from very large data bases generalized by dynamic conceptual hierarchies provided by users. In general, the process of attribute generalization may introduce inconsistency into a generalized relation. This issue is resolved by using the inductive learning algorithm, LERS based on rough set theory.

Determination of quantization intervals in rule based model for dynamic systems

The authors introduce two adaptive procedures for quantizing continuous data used by symbolic empirical learning programs to generate rule-based models for dynamic systems. The basic idea is to use a top-down iterative procedure for refining quantization intervals selectively. In each iteration, the quantization interval having a maximum overall error rate is selected for refining. Each time a selected interval is divided into two new equal intervals. Based on the new quantization intervals, a new set of rules is generated and performance associated with each quantization interval is evaluated again. The refining procedure is applied repeatedly until a user-specified performance index is reached. The method was tested by two examples, one involving a simulated system, and the other a real life gas furnace.<<ETX>>

Transactions on rough sets XIII

Read more and get great! Thats what the book enPDFd transactions on rough sets xiii will give for every reader to read this book. This is an on-line book provided in this website. Even this book becomes a choice of someone to read, many in the world also loves it so much. As what we talk, when you read more every page of this transactions on rough sets xiii, what you will obtain is something great.

SVM Approach to Breast Cancer Classification

Inference of evolutionary trees using the maximum likelihood principle is NP-hard. Therefore, all practical methods rely on heuristics. The topological transformations often used in heuristics are nearest neighbor interchange (NNI), sub-tree prune and regraft (SPR) and tree bisection and reconnection (TBR). However, these topological transformations often fall easily into local optima, since there are not many trees accessible in one step from any given tree. Another more exhaustive topological transformation is p-Edge Contraction and Refinement (p-ECR). However, due to its high computation complexity, p-ECR has rarely been used in practice. This paper proposes a method p-ECRNJ with a O(p3) time complexity to make the p-ECR move efficient by using neighbor joining (NJ) to refine the unresolved nodes produced in p-ECR. Moreover, the demonstrated topological accuracy for small datasets of NJ can guarantee the accuracy of the p-ECRNJ move. Experiments with simulated and real datasets show that p-ECRNJ can find better trees than the best-known maximum likelihood methods so far and can efficiently improve local topological transforms in reasonable time.The purpose of the proposed study was to provide a solution to the Wisconsin diagnostic breast cancer (WDBC) classification problem. The WDBC dataset, provided by the University of Wisconsin Hospital, was derived from a group of images using fine needle aspiration (biopsies) of the breast. An ensemble of support vector machines (SVMs) was employed in this study. Support vectors with linear, polynomial and RBF kernel functions were trained using a fraction of the WDBC dataset as a training set. The five top performing models were recruited into the ensemble. The classification was then carried out using the majority opinion of the ensemble. The SVM ensemble successfully classified more than 99 percent of the testing data and in the process yielded 100 percent benign tumor prediction accuracy.

Automated rule based model generation for uncertain complex dynamic systems

An attempt is made to model a dynamic system by a set of production rules. These rules are automatically induced from a set of training data by the ID3 algorithm of J.R. Quinlan (1983, 1986). The effects of data quantization and number of attributes on the model performance are discussed. The algorithm is applied to a simulated linear system and to real gas furnace data.<<ETX>>

Support vector machines for fault detection

Support vector machines (SVMs), based on Vapniks statistical learning theory is a new tool that can be used for fault detection and isolation in dynamic systems. This paper presents a new approach that combines the system identification technique and the SVM learning algorithm for fault detection and isolation in dynamic systems. A conventional heat exchanger dynamics is used to illustrate the technique.

Combined rough sets with flow graph and formal concept analysis for business aviation decision-making

Although business aviation has been popular in the USA, Europe, and South America, however, top economies in East Asia, including Japan, Korea, and Taiwan, have been more conservative and lag behind in the development of business aviation. In this paper, we hope to discover possible trends and needs of business aviation for supporting the government to make decision in anticipation of eventual deregulation in the near future. We adopt knowledge-discovery tools based on rough set to analyze the potential for business aviation through an empirical study. Although our empirical study uses data from Taiwan, we are optimistic that our proposed method can be similarly applied in other countries to help governments there make decisions about a deregulated market in the future.

Mining pharmaceutical spam from Twitter

This paper presents a method of applying text mining techniques and data mining tools for pharmaceutical spam detection from Twitter data. A simple method based on a manually selected list of 65 pharmaceutical discriminating words is used for labeling spam training tweets. Preliminary experimental results show that J48 decision tree classifier has better performance over Naïve Bayesian algorithm.

Web usage mining using rough sets

This paper studies the use of a rough set based learning program for predicting Web usage. In our approach, Web usage patterns are represented as rules generated by the inductive learning program, BLEM2. Inputs to BLEM2 are clusters generated by a hierarchical clustering algorithm applied to preprocessed Web log records. Empirical results show that the prediction accuracy of rules induced by the learning program is better than a centroid based method. In addition, the use of a learning program can generate shorter cluster descriptions.

Using inductive learning to determine fuzzy rules for dynamic systems

Abstract This paper describes a fuzzy model-building methodology. Process input-output data are quantized by fuzzy reference membership functions. Membership values are determined for each data point and only the membership function that generates the maximum membership value is used. The name of membership functions corresponding to maximum values are fed into an inductive learning algorithm and the fuzzy rules are determined. The methodology is applied to fuzzy modeling of a dynamic system.