Article View: aus.computers.ai
Article #861Vic AISIG seminar 19/8: machine learning, Wayne Jenkins
From: Ross Gayler
Date: Tue, 13 Aug 1996 00:00
Date: Tue, 13 Aug 1996 00:00
84 lines
3263 bytes
3263 bytes
ACS AISIG (Vic)
Public Seminar
5:45pm - 7:30pm
Monday 19th August
Basement Lecture Theatre 2
Computer Science Department., Melbourne University
221 Bouverie Street, Carlton
***** Everyone is welcome to attend our seminars and come
out for a meal afterwards with the speaker. *****
SPEAKER
Wayne Jenkins
Department of Computer Science, RMIT
TITLE
Myths and Legends: A perspective on machine learning
ABSTRACT
Machine learning practitioners make use of statistics to
show that their learning algorithms are performing well.
Such statistics, however, are often used in a way that
misleads the reader as to the significance of the result
of a particular study. In this presentation we expose
some of the myths that seem to be prevalent amongst some
practitioners of machine learning both in the use of
statistics and the implications that can be drawn from
these statistics. We discuss the no-free-lunch (or
conservation law) theorems, the implications of such
theorems and (to be controversial) we point out why
neural networks may not be the answer. We conclude by
considering current approaches to machine learning. In
particular, we discuss the speakers own research into
finding learning algorithms which, under certain
circumstances, are optimal.
This presentation has been prepared with the intention of
appealing to a broad audience. Esoterics will be kept to
a minimum.
ABOUT THE SPEAKER
Wayne Jenkins is a lecturer in computer science at RMIT
who has recently completed his PhD thesis on machine
learning.
MEETING LOCATION
Our meeting venue is Lecture Theatre 2, Basement Level,
221 Bouverie Street, Carlton. This is the Computer
Science Building of the University of Melbourne. Hints
on getting there:
1. The stairs to the basement are in the north-east
corner of the building. Do not use the doors in the
centre of the building front.
2. If tramming: Get off the Swanston Street tram at
Lincoln Square then walk north-west across the park.
Look for the five story grey-green building.
3. If driving: The building is 100 metres south of the
T-intersection with Grattan Street (the street that runs
past the southern border of Melbourne University. The
parking can be tricky during term time, so make sure you
get there a few minutes early to give you time to look
around. Usually, parking spots can be found reasonably
close.
FURTHER INFORMATION
If you have any questions, please call:
Ross Gayler 9643 8853
Simon Goss 9626 7274
Julie Gledhilll 9818 0795
Technical Details
Message-ID:
<4upqiv$5oh@hobyah.cc.uq.oz.au>Path:
arc1-ca.pugleaf.net!unknown!not-for-mailShow Full Headers
Complete Article Headers
X-Google-Language: ENGLISH,ASCII-7-bit X-Google-Thread: 1132b9,de2206561e6f203a,start X-Google-Attributes: gid1132b9,public From: ross@psy.uq.oz.au (Ross Gayler) Subject: Vic AISIG seminar 19/8: machine learning, Wayne Jenkins Date: 1996/08/13 Message-ID: <4upqiv$5oh@hobyah.cc.uq.oz.au> X-Deja-AN: 173898353 distribution: aus organization: University of Queensland newsgroups: aus.computers.ai