Fwd: multivariate ranks clustering (a.k.a. "An Olympic question")

---------- Forwarded message ----------
From: Michael GREENACRE <michael.greenacre@upf.edu>
Date: 2012/8/5
Subject: Re: multivariate ranks clustering (a.k.a. "An Olympic question")
To: Basilio de Bragança Pereira <basilio@hucff.ufrj.br>

Hi Basilio

I'm not sure about this, but did you hear my SVD song on youtube.com/StatisticalSongs?:

http://youtu.be/JEYLfIVvR9I

Michael

De Basilio de Bragança Pereira <basilio@HUCFF.UFRJ.BR>

Enviado Mié 25/07/2012 22:07

Para allstat@JISCMAIL.AC.UK

Asunto Re: multivariate ranks clustering (a.k.a. "An Olympic question")

Hi Fabio
This is Basilio from Rio de Janeiro
Have you tried SVD in thsi matrix to see if you find any pattern?
If you have matrices of Arrival , Scores ,etc...try single SVD of the tensors resulting . See Understanding Complex Data Sets :Data Mining with Matrix Decomposition (Skillicorn Chapman Hall)
.
Best regards
Basilio

Hi Fabio
This is Basilio from Rio de Janeiro
Have you tried SVD in thsi matrix to see if you find any pattern?
If you have matrices of Arrival , scores ,...try single SVD of the tensors resulting . See Understanding Complex Data Sets :Data Mining with Matrix Decomposition (Skillicorn Chapman Hall)
.
Best regards
Basilio

2012/7/25 Rigat, Fabio <fabio.rigat@novartis.com>

Dear all,

Recommendations are welcome on published methods to address the following type of question:

A fixed set of athletes 1:N participate to 1:p distinct competitions. From the full N times p matrix of results (e.g. Arrival times, scores,…) an equivalent size matrix of ranks is calculated.

Column 1:p of this rank matrix lists a 1 for first classified and so forth up to N for the last athlete for each competition.

A sports commentator wishes to cluster this matrix so as to discriminate groups of athletes associated to a distinctive pattern of results.

For instance, some athletes may be more versed for endurance competitions and less so for short-distance fast competitions.

A priori, the number and nature of these patterns are unknown.

Best wishes

Fabio Rigat

You may leave the list at any time by sending the command
SIGNOFF allstat
to listserv@jiscmail.ac.uk, leaving the subject line blank.

Basilio de Bragança Pereira ,DIC and PhD(Imperial College), DL(COPPE)
*UFRJ-Federal University of Rio de Janeiro
*Titular Professor of Bioestatistics and of Applied Statistics
*FM-School of Medicine and COPPE-Posgraduate School of Engineering and
HUCFF-University Hospital Clementino Fraga Filho.

*Tel: 55 21 2562-7045/7047/2618/2558
www.po.ufrj.br/basilio/

*MailAddress:
COPPE/UFRJ
Caixa Postal 68507
CEP 21941-972 Rio de Janeiro,RJ
Brazil

You may leave the list at any time by sending the command

SIGNOFF allstat

to listserv@jiscmail.ac.uk, leaving the subject line blank.

*MailAddress:
COPPE/UFRJ
Caixa Postal 68507
CEP 21941-972 Rio de Janeiro,RJ
Brazil