How One Can Estimate Web-Access Rates with (Almost) No Data

Speaker:	Prof. Norm Matloff
		Department of Computer Science
		University of California, Davis
		USA

Topic:		How One Can Estimate Web-Access Rates with
		(Almost) No Data

Date:		Monday, 22 November 2004

Time:		4:00 pm - 5:00 pm

Venue:		Lecture Theatre F
		(Leung Yat Sing Lecture Theatre,near lift nos. 25/26)
		HKUST

Abstract:

Suppose we wish to estimate access/modification (a/m) rates of Web pages.
Data is collected at periodic intervals, but it is not complete a/m data.
All that is available is time of the MOST RECENT a/m access in each
interval.  Can one estimate a/m rates from this kind of indirect data?
Surprisingly, the answer is yes.  Here we develop methodology for such
estimation.  Both parametric and nonparametric methods are developed.
Mathematical and empirical analyses are presented which indicate that the
problem is indeed statistically tractable, and that the methods developed
are of practical value.


**********************
Biography:

Dr. Norm Matloff is a professor of computer science at the University of
California, Davis.  He is a former professor of statistics at UCD,  and he
is also a former software developer in industry.  Dr. Matloff has served
as an appointed member of IFIP Working Group 11.3, an international
committee concerned with statistical database security, and his work on
optical multiprocessor computers was awarded a U.S. patent.  He is the
author of KuaiXue, a software tool for learning written Chinese, and
PerlDSM, a system which enables parallel programming in Perl.  His
research interests are in computer systems, parallel processing, data
security and data mining.