scispace - formally typeset
Search or ask a question

Showing papers presented at "International Conference on Bioinformatics in 2003"


Proceedings Article
01 Jan 2003

2 citations


Proceedings Article
01 Jan 2003

1 citations


Journal ArticleDOI
30 Jun 2003
TL;DR: A variety of lexical and link based approaches of focused web crawling are introduced in this article, highlighting important aspects of each of them, and the need for more and more specific reply to a web search query has prompted researchers to work on focused Web crawling techniques for web spiders.
Abstract: The need for more and more specific reply to a web search query has prompted researchers to work on focused web crawling techniques for web spiders. Variety of lexical and link based approaches of focused web crawling are introduced in the paper highlighting important aspects of each.

1 citations


Proceedings Article
01 Jan 2003
TL;DR: Mass-radius and wavelet multifractal analysis are most suitable for investigating neurons for multifractality and the application of unsupervised learning algorithms combined with additional feature parameters in order to investigate differences between post mortem tissue of people with alcoholism and controls is discussed.
Abstract: Neuroinformatics provides a platform to share information on a previously unimaginable grand scale. It provides the opportunity to discover new relationships between structure and function of neurons by utilizing large data sets consisting of existing and novel computational analysis results. Fractal analysis of the morphology of neurons has been reported in the literature for nearly 15 years [2]. In that time advances in the mathematics of fractals have suggested that biological objects may be multifractal rather than monofractal. However due to the finite size of biological objects, box-counting has not been able to answer this question. In this paper we discuss advances in multifractal analysis using mass-radius and wavelet methodology and combining this with unsupervised learning classification algorithms using human post mortem material. Moreover we apply some already very known tools [1] to the image under investigation. Some preliminary results indicate that human precentral and superior frontal gyrus layer III pyramidal cells have multifractal features. Furthermore, the analysis differentiates between the regions. Mass-radius and wavelet multifractal analysis are most suitable for investigating neurons for multifractality. Our ongoing project includes the application of unsupervised learning algorithms combined with additional feature parameters[3] in order to investigate differences between post mortem tissue of people with alcoholism and controls. REFERENCES [1]Cesar, Jr.,R. M., Leandro, J. J. G. and Jelinek, H. F., Exploring the structure-function relationship of cat retinal ganglion cells using wavelets, Brain & Mind, 2003 (in press). [2]Jones, C. L., Jelinek, H. F. , Wavelet Packet Fractal Analysis of Neuronal Morphology, August 2001, Vol 24, No. 4, 347-358 [3]Loncaric, S. A Survey of Shape Analysis Techniques, Pattern Recognition, Vol 31(8) (1998) 983-1001

1 citations