Rongfang Bie

Automatic web pages categorization with ReliefF and Hidden Naive Bayes

A great challenge of web mining arises from the increasingly large web pages and the high dimensionality associated with natural language. Since classifying web pages of an interesting class is often the first step of mining the web, web page categorization/classification is one of the essential techniques for web mining. One of the main challenges of web page classification is the high dimensional text vocabulary space. In this research, we propose a Hidden Naive Bayes based method for web page classification. We also propose to use the ReliefF feature selection method for selecting relevant words to improve the classification performance. Comparisons with traditional techniques are provided. Results on benchmark dataset show that the proposed methods are promising for accurate web page classification.

Artificial Immune Networks: Models and Applications

In this paper, the theory of natural immune system is first briefly introduced. Several representative artificial immune networks are next discussed. Their principles and learning algorithms are given here in details. Moreover, we demonstrate the applications of these artificial immune networks in the fields of data mining, pattern recognition, and optimization

Spam email filtering with bayesian belief network: using relevant words

In this paper, we report our work on a Bayesian Belief Network approach to spam email filtering (classifying email as spam or nonspam/legitimate). Our evaluation suggests that a Bayesian Belief Network based classifier will outperform the popular Naive Bayes approach and two other famous learners: decision tree and k-NN. These four algorithms are tested on two different data sets with three different feature selection methods (Information Gain, Gain Ratio and Chi Squared) for finding relevant words. 10-fold cross-validation results show that Bayesian Belief Network performs best on both datasets. We suggest that this is because the dependant learner characteristics of Bayesian Belief Network classification are more suited to spam filtering. The performance of the Bayesian Belief Network classifier could be further improved by careful feature subset selection.

A new clustering algorithm based on artificial immune network and K-means method

This paper proposes a new data clustering method, which is based on artificial immune network and k-means method. With a pool of memory cells, the artificial immune network can be used for estimating the input data distribution, while the k-means method has the capability of shaping clear clusters and obtaining their centers. On the basis of an improved artificial immune network, we first cluster the memory cells by using the k-means algorithm, and with the generated data clusters, we can make the data classification or prediction. The results of our experiments on the standard data sets demonstrate that this new algorithm has a superior performance of data clustering and classification.

Artificial Immune Recognition System for DNA Microarray Data Analysis

Artificial immune systems (AIS) are emerging information processing methods, which embody the principles of biological immune systems for tackling complex realworld problems. The artificial immune recognition system (AIRS) is a new kind of supervised learning AIS. The development of microarray technology has supplied a large volume of data for the prediction and diagnosis of cancer. Many popular machine learning techniques have been used in the microarray data analysis. In this paper, we apply AIRS to perform the microarray data classification based on an improved version of the information gain feature selection method. Three traditional classifiers have also been employed in our experiments for performance comparison. The results demonstrate the promising ability of AIRS in the microarray data analysis.

BP Neural Networks with Harmony Search Method-based Training for Epileptic EEG Signal Classification

In this paper, the Harmony Search (HS)-based BP neural networks are used for the classification of the epileptic electroencephalogram (EEG) signals. It is well known that the gradient descent-based learning method can result in local optima in the training of BP neural networks, which may significantly affect their approximation performances. Two HS methods, the original version and a new variation recently proposed by the authors of the present paper, are applied here to optimize the weights in the BP neural networks for the classification of the epileptic EEG signals. Simulations have demonstrated that the classification accuracy of the BP neural networks can be remarkably improved by the HS method-based training.

Cancer classification from serial analysis of gene expression with event models

AbstractnCancer class prediction and discovery is beneficial to imperfect non-automated cancer diagnoses which affect patient cancer treatments. Serial Analysis of Gene Expression (SAGE) is a relatively new method for monitoring gene expression levels and is expected to contribute significantly to the progress in cancer treatment by enabling an automatic, precise and early diagnosis. Axa0promising application of SAGE gene expression data is classification of cancers. In this paper, we build three event models (the multivariate Bernoulli model, the multinomial model and the normalized multinomial model) for SAGE gene expression profiles. The event models based methods are compared with the standard Naïve Bayes method. Both binary classification and multicategory classification are investigated. Experiments results on several SAGE datasets show that event models are better than standard Naïve Bayes in general. Normalized Information Gain (NIG), an extension of Information Gain (IG), is proposed for gene selection. The impact of gene correlation on the classification performance is investigated.n

A Comparison Study of Bayesian Classifiers on Web Pages Classification

With the development of internet, web mining has become a hotspot of data mining. The first step of web mining is to classify web pages into interesting classes, so the classification is one of the essential techniques for web mining. In this paper, we study the capabilities of bayesian classifiers for web pages categorization, after that we report our work on the comparison of binary-classification and multi-classification. Results on benchmark dataset show that bayesian classifiers perform satisfying, especially for Hidden Naive Bayes (HNB) which is more competitive than other methods. Furthermore, the performances of binary-classification are better than those of multi-classification under the evaluation metrics of accuracy and F-measure.

Global and local preserving feature extraction for image categorization

In this paper, we describe a feature extraction method: Global and Local Preserving Projection (GLPP). GLPP is based on PCA and the recently proposed Locality Preserving Projection (LPP) method. LPP can preserve local information, while GLPP can preserve both global and local information. In this paper we investigate the potential of using GLPP for image categorization. More specifically, we experiment on palmprint images. Palmprint image has been attracting more and more attentions in the image categorization/recognition area in recent years. Experiment is based on benchmark dataset PolyU, using Error Rate as performance measure. Comparison with LPP and traditional algorithms show that GLPP is promising.

MSC: A Semantic Ranking for Hitting Results of Matchmaking of Services

As the e-commerce is done faster, there is a continuous flourishing of e-marketplaces. Matchmaking is an important aspect of e-commerce interactions. Recently, an approach has been taken to service matchmaking based on semantic Web technologies; the designed matching rule can be used to find the sellers compatible advertisements for buyers. In this paper, we define three categories of attributes for the matchmaking service. Then we present three factors: semantic matching degree, semantic support and relational confidence to capture the semantic characteristics and relationships of the attributes. And we design a semantic ranking MSC combining the three factors to rank the results of advertisements matchmaking. MSC can capture the semantic aspect of matchmaking results; evaluation shows that MSC makes the process of matchmaking more accurately and the advertisement with the highest MSC is better