Vincent T. Y. Ng

Efficient mining of association rules in distributed databases

Many sequential algorithms have been proposed for the mining of association rules. However, very little work has been done in mining association rules in distributed databases. A direct application of sequential algorithms to distributed databases is not effective, because it requires a large amount of communication overhead. In this study, an efficient algorithm called DMA (Distributed Mining of Association rules), is proposed. It generates a small number of candidate sets and requires only O(n) messages for support-count exchange for each candidate set, where n is the number of sites in a distributed database. The algorithm has been implemented on an experimental testbed, and its performance is studied. The results show that DMA has superior performance, when compared with the direct application of a popular sequential algorithm, in distributed databases.

Dullrazor®: A software approach to hair removal from images

Recently, there has been a growing number of studies applying image processing techniques to analyze melanocytic lesions for atypia and possible malignancy and for total-body mole mapping. However, such lesions can be partially obscured by body hairs. None of these studies has fully addressed the problem of human hairs occluding the imaged lesions. In our previous study we designed an automatic segmentation program to differentiate skin lesions from the normal healthy skin, and learned that the program performed well with most of the images, the exception being those with hairs, especially dark thick hairs, covering part of the lesions. These thick dark hairs confused the program, resulting in unsatisfactory segmentation results. In this paper, we present a method to remove hairs from an image using a pre-processing program we have called DullRazor. This pre-processing step enables the segmentation program to achieve satisfactory results. DullRazor can be downloaded as shareware from http:/(/)www.derm.ubc.ca.

Molecular Pattern Discovery Based on Penalized Matrix Decomposition

A reliable and precise identification of the type of tumors is crucial to the effective treatment of cancer. With the rapid development of microarray technologies, tumor clustering based on gene expression data is becoming a powerful approach to cancer class discovery. In this paper, we apply the penalized matrix decomposition (PMD) to gene expression data to extract metasamples for clustering. The extracted metasamples capture the inherent structures of samples belong to the same class. At the same time, the PMD factors of a sample over the metasamples can be used as its class indicator in return. Compared with the conventional methods such as hierarchical clustering (HC), self-organizing maps (SOM), affinity propagation (AP) and nonnegative matrix factorization (NMF), the proposed method can identify the samples with complex classes. Moreover, the factor of PMD can be used as an index to determine the cluster number. The proposed method provides a reasonable explanation of the inconsistent classifications made by the conventional methods. In addition, it is able to discover the modules in gene expression data of conterminous developmental stages. Experiments on two representative problems show that the proposed PMD-based method is very promising to discover biological phenotypes.

Evolutionary time series segmentation for stock data mining

Stock data in the form of multiple time series are difficult to process, analyze and mine. However, when they can be transformed into meaningful symbols like technical patterns, it becomes easier. Most recent work on time series queries concentrates only on how to identify a given pattern from a time series. Researchers do not consider the problem of identifying a suitable set of time points for segmenting the time series in accordance with a given set of pattern templates (e.g., a set of technical patterns for stock analysis). On the other hand, using fixed length segmentation is a primitive approach to this problem; hence, a dynamic approach (with high controllability) is preferred so that the time series can be segmented flexibly and effectively according to the needs of users and applications. In view of the fact that such a segmentation problem is an optimization problem and evolutionary computation is an appropriate tool to solve it, we propose an evolutionary time series segmentation algorithm. This approach allows a sizeable set of stock patterns to be generated for mining or query. In addition, defining the similarity between time series (or time series segments) is of fundamental importance in fitness computation. By identifying perceptually important points directly from the time domain, time series segments and templates of different lengths can be compared and intuitive pattern matching can be carried out in an effective and efficient manner. Encouraging experimental results are reported from tests that segment the time series of selected Hong Kong stocks.

Collaborative solid modeling on the WWW

This project studied some of the issues associated with collaborative solid modeling (CSM) on the WWW and implemented some possible solutions. The CSM system developed oglers the following facilities for collaborative design: (I) an environment for multiple users to edit a shared solid object on the WWW synchronous!v, (2) representation of alternate versions of a solid by a single CSG tree, (3) support for dif/rerent modes of collaboration among the users. The system was written in JAVA and tested mainly in a UNIX environment.

A comparative study of biweekly disposable contact lenses: silicone hydrogel versus hydrogel.

Background:  Our aim was to compare the clinical performance of a biweekly (second generation) silicone hydrogel lens and a biweekly hydrogel lens worn for daily wear modality.

A multi-stage segmentation method for images of skin lesions

We present an algorithm to identify skin lesions from the digitized colour images collected in a clinical study of malignant melanomas in early 1994. The algorithm consists of three steps: (1) a multi-stage median filter to suppress noise, (2) a process to compute the threshold values for the lesions and the background normal skin, and (3) a rule-based system to identify the lesions. Trying to make full use of the colour information, the first two steps of the algorithm work on the RGB planes. Then the three colour planes are re-combined before the lesions are identified in the last step.

Variability of tear protein levels in normal young adults: between-day variation

Abstract Background: An adequate knowledge of physiological variation is important for valid comparative studies of tear proteins. The aim of this study was to investigate the between-day variation of the human tear protein levels, including the total protein concentration (TPC) and the levels of major protein fractions. Two sampling methods, the yawn and the eye-flush methods, were used and compared. Methods: TPC was determined by the Bradford method. The major protein fractions were separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) and their levels were determined by scanning densitometry after SDS-PAGE. The tear protein levels were monitored for 3 days. Results: The between-day differences in the levels of TPC and the individual protein fractions were not statistically significant in either sampling method, but the variations of some proteins were large and would be clinically significant. Different variations were observed in different proteins. The variations in serum albumin were large, up to 61% and 113% in the yawn and eye-flush methods respectively. The variations in lactoferrin, tear-specific prealbumin and lysozyme were relatively small in the yawn method. The variations in protein levels obtained by the eye-flush method were generally higher than by the yawn method. Conclusion: Although the between-day differences in tear protein levels were not statistically significant, the variations in some proteins would be large in magnitude. The variability of tear protein levels obtained by the eye-flush method was larger than that by the yawn method. Therefore, caution should be taken if the eye-flush method is used for sampling tears for quantitative analysis of tear proteins, although it is easier to collect tear samples using this method. The results will be useful to exclude normal variation in tear protein levels when comparing pre- and post-therapeutic tear protein levels in eyes treated for tear-related abnormalities.

Tear proteins of normal young Hong Kong Chinese

Abstract Background: Analysis of tear proteins is of diagnostic value for abnormal ocular conditions such as dry eye syndrome. Many studies of tear proteins have been performed on Caucasian subjects. However, little is known about these proteins in Chinese eyes. Methods: The total tear protein concentrations of 30 normal young Hong Kong Chinese were determined by the Bradford method and the modified Lowry method. Bovine serum albumin (BSA) and bovine immunoglobulin G (IgG) were both used as standards for each method. The tear protein patterns were determined by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE), and the concentrations of major tear proteins were quantified by scanning densitometry after SDS-PAGE. Results: The mean±SD total tear protein concentrations determined by the Bradford method, using BSA and IgG as standards, were 6.05±1.58 mg/ml and 11.48± 2.32 mg/ml respectively. The values determined by the modified Lowry method, using the same two standards, were 9.66±2.03 mg/ml and 7.53±1.80 mg/ml respectively. The mean±SD concentrations of major tear proteins were 2.73±0.82 mg/ml for lactoferrin, 0.021±0.028 mg/ml for human serum albumin, 2.89± 0.88 mg/ml for tear-specific prealbumin and 2.46±0.44 mg/ml for lysozyme. Conclusion: The results of total tear protein concentrations indicated that values obtained from different methods and different standards were not comparable. The tear protein patterns of our subjects were qualitatively similar to those reported for Caucasian subjects. However, the concentrations of the major proteins of our subjects were not in accordance with those reported previously. The main reason may be the large variability of method used.

Evolutionary segmentation of financial time series into subsequences

Time series data are difficult to manipulate. When they can be transformed into meaningful symbols, it becomes an easy task to query and understand them. While most recent works in time series query only concentrate on how to identify a given pattern from a time series, they do not consider the problem of identifying a suitable set of time points based upon which the time series can be segmented in accordance with a given set of pattern templates, e.g., a set of technical analysis patterns for stock analysis. On the other hand, using fixed length segmentation is only a primitive approach to such kind of problem and hence a dynamic approach is preferred so that the time series can be segmented flexibly and effectively. In view of the fact that such a segmentation problem is actually an optimization problem and evolutionary computation is an appropriate tool to solve it, we propose an evolutionary segmentation algorithm in this paper. Encouraging experimental results in segmenting the Hong Kong Hang Seng Index using 22 technical analysis patterns are reported.