This paper proposes a new algorithm to classify pulse waveforms based on discrete wavelet network. This paper selects 4-order discrete Daubechies wavelet as the wavelet node of this wavelet network to classify six pulse patterns distinctive in shape. 600 pulse records are used to train this wavelet network and 300 pulse records are used to test the classifier's performance. The test results show that this approach has 83% agreement rate with the experienced experts. Compared with traditional classification methods, it needs not the experience in feature extraction.

A numerical investigation on the effects of small tripping rods on the fluid force reduction on a big structure has been carried out using finite volume method where a configuration of a circular cylinder with two small tripping rods symmetrically placed very near to its front surface is studied. The diameter ratio of the rods and the cylinder is set at 0.08,0.10 and 0.12, and the gap between the rods and the cylinder is fixed at 0.08 of the cylinder diameter. The angular position of the rods varies from 20°to 60°. The effects of the tripping rods on force reduction, vortex shedding frequency and flow separation have been examined for various arrangements of the rods with Reynolds number focused on 200 for laminar flow and 5.5xl0 4 for a turbulent flow. The results reveal that there exits an optimum position where the time averaged force coefficients acting on the cylinder all reach their minimum values and at the same time Strouhal number meets its maximum. At the optimum position the drag coefficient is reduced by 18% for Re=200 and 59% for Re=5.5xI0 4 • Further investigation with tripping rods placed near the separation points is also carried out for Re=5.5x 10 4 and a considerable drag reduction is found.

Abstract

The computerization of traditional Chinese pulse diagnosis (TCPD) is relatively new in the field of automatic physiological signal analysis and diagnosis. The classification of pulse patterns according to their positions and shapes have been intensively investigated, but until now no research in identifying pulses by their rhythms has been found. This paper introduces a method to distinguish rhythmic and arrhythmic pulse patterns and further, applies the Lempel-Ziv decomposition to classify arrhythmic pulses. In the experiment on 140 clinic pulses, our approach was able to classify pulses by their rhythms with an accuracy of 90.7%.

DNA microarray experiments generate a substantial amount of information about global gene expression. Gene expression profiles can be represented as points in multi-dimensional space. It is essential to identify relevant groups of genes in biomedical research. Clustering is helpful in pattern recognition in gene expression profiles. Some clustering techniques have been introduced. However, these traditional methods mainly utilize shape-based assumption or distance metric to cluster the points in multi-dimension linear Euclidean space. Poor consistence with the functional annotation of genes is shown in their validation study.We propose fractal clustering method to cluster genes using intrinsic (fractal) dimension from modern geometry. Fractal dimension is used to characterize the degree of self-similarity among the points in the clusters. The main idea of fractal clustering is to group points in a cluster in such a way that none of the points in the cluster changes the cluster's intrinsic dimension radically. We computed Hausdorff fractal dimension through the means of the box-counting plot algorithm, since it is the fastest and also robust enough.We assess this method using validation assessment using public microarray dataset. It shows that this method is superior in identifying functional related gene groups than other traditional methods.