[ASC-media] ECU Natural Sciences July 27 & 29: data analysis and Open Day!

Nardia BORDAS n.bordas at ecu.edu.au
Mon Jul 23 11:00:03 CEST 2012

Greetings, Science Fans!

Apologies for the late notice of last week's seminar, we are endeavouring to rectify the glitches.  In the meantime we have two events coming up: another seminar and our campus Open Day.

You are invited to our next SoNS (School of Natural Sciences) seminar for 2012 at 3.30pm on Friday 27th July. Dr Raimon Tolosana Delgado from the Technical University of Catalonia will discuss compositional data analysis using environmental case studies (see abstract below).
These are not "dumbed down" talks but real research presentations so content may be quite technical at times; nevertheless all interested members of the public are welcome to join us for the seminar plus drinks and nibbles afterwards, where you can chat with the researchers in person.
Places are strictly limited so RSVPs are essential.  To reserve your place, receive parking and venue details or to request further information, e-mail n.bordas at ecu.edu.au<mailto:n.bordas at ecu.edu.au> or leave a message on (08) 6304 5116.
If you'd like the maximum notice of upcoming seminars, reply to this e-mail asking to be put in contact with the series convenor.

Your best chance all year to see our facilities in action is at our campus Open Day, this Sunday July 29 from 10am to 4pm.  We'll have an open lab, a Chemistry Show, academic staff to ask about our courses and much, much more!  Visit http://www.ecu.edu.au/future-students/?a=74893 to see what's on or schedule your day with our online planner.

In science,
Nardia Bordas
Promotions Officer
School of Natural Sciences, Edith Cowan University
Room JO19.242, 270 Joondalup Drive, Joondalup WA 6027
T +61 8 6304 5116, F +61 8 6304 5842
n.bordas at ecu.edu.au<mailto:n.bordas at ecu.edu.au>


 Friday 27th July 3:30pm
Building 19 (Room 19.141) Joondalup

Dr Raimon Tolosana Delgado
Laboratory of Maritime Engineering (LIM-UPC), Technical University of Catalonia

An introduction to Compositional Data Analysis

Compositional data are typically sets of variables expressed in percentages, ppm or in proportions. They appear in virtually any field of science, where one wants to express the relative importance of several parts forming a whole: weight % of major oxides in a soil or a rock; proportion of certain taxa that have fallen in an insect trap; concentration of several ions in water;  % of votes in an election going to each party. It has been recognized that conventional correlation analysis applied to this kind of data tend to be negatively biased, and often provide spurious results (and the same happens with many covariance-based statistical tools). To avoid such problems, we must first realize that compositional data cannot tell us anything about absolute increments or decrements of our components. Actually, only (log-)ratios between components are fully meaningful. Compositional Data Analysis starts from these realizations to build a set of tools devised to work around the spurious correlation problems. This talk will present these principles and tools, illustrated with two case studies: elucidation of pollution sources in the Llobregat River Basin running waters (Barcelona, Spain) out of a dataset of ionic concentrations; and the mapping of percentages of soil components for a geochemical Atlas of the Kola Peninsula (at the border between Finland, Russia and Norway).

Refreshments in SONS staff room, Bldg 19 following.
All welcome.

This e-mail is confidential. If you are not the intended recipient you must not disclose or use the information contained within. If you have received it in error please return it to the sender via reply e-mail and delete any record of it from your system. The information contained within is not the opinion of Edith Cowan University in general and the University accepts no liability for the accuracy of the information provided.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.asc.asn.au/pipermail/asc-media/attachments/20120723/eeae8027/attachment.html>

More information about the ASC-media mailing list