LIEF progress report 2004: UQ
- Contemporary messages sorted: [ by date ] [ by thread ] [ by subject ] [ by author ] [ by messages with attachments ]
From: <mjd_at_email.protected>
Date: Wed, 2 Feb 2005 08:33:31 +1000 (EST)
Dear colleagues,
Some of you (especially Rachel) may need progress report material for the
ARC, so here is a current summary of what we did at UQ on the 2004 LIEF
grant. Note that we started in March 2004, so our main deliverable is on
track for March 2005.
Dr. Michael Drinkwater tel: 07 3365 3428
Department of Physics fax: 07 3365 1242
University of Queensland email: <a href="mailto:mjd_at_physics.<!--nospam-->uq.edu.au
_at_physics.uq.edu.au?Subject=Re:%20LIEF%20progress%20report%202004:%20UQ">mjd_at_physics.uq.edu.au</a>
QLD 4072, Australia http://www.physics.uq.edu.au/ap
---------- Forwarded message ----------
Background:
At the University of Queensland a cross-disciplinary team has focused
on the problem that positional uncertainty poses in terms of matching
catalogues. In 2003 (funded by an internal UQ grant) we were able to show
that machine learning techniques could be used to match the SuperCOSMOS
and HIPASS catalogues using a model we trained on a linked subset.
Personnel:
Dr. M. Drinkwater (Physics UQ)
Prof T. Downs, Dr. M. Gallagner (ITEE, UQ)
Mr. D. Rohde (Physics and ITEE, UQ)
Activities 2004:
1. We have developed a prototype system that can resolve the ambiguous
matching problem for arbitrary catalogues using a user-provided training
set. The prototype service has been installed on a 16-node
high-performance computing cluster located at UQ. [The batch server
component is now being installed; we will make this available to the IVO
group by March 2005.]
2. David Rohde, who built the prototype system, attended The US National
Virtual Observatory Summer School (Sep 2004) in order for the team to gain
greater understanding of the world context of Virtual Observatory planning
and software. His team won a competition held during the Summer School to
build a VO demonstration tool.
3. The development work has resulted in these publications/presentations:
Refereed:
`Machine Learning for Matching Catalogues', Rohde D, Drinkwater M,
Gallagher M, Downs T, Doyle M, 2004, Proceedings of Intelligent Data
Engineering and Automated Learning, 1.3177, 702-707
`Applying Machine Learning to Catalogue Matching in Astrophysics', Rohde
D, Drinkwater M, Gallagher M, Downs T, Doyle M, 2005, MNRAS, submitted
Conferences:
Intelligent Data Engineering and Automated Learning Conference 2004
`Machine Learning for Matching Catalogues'
Annual Scientific Meeting of the Astronomical Society of Australia 2004
'Machine Learning for Matching Catalogues'
###########################################################################
Received on 2005-02-02 09:34:02
Date: Wed, 2 Feb 2005 08:33:31 +1000 (EST)
Dear colleagues,
Some of you (especially Rachel) may need progress report material for the
ARC, so here is a current summary of what we did at UQ on the 2004 LIEF
grant. Note that we started in March 2004, so our main deliverable is on
track for March 2005.
Dr. Michael Drinkwater tel: 07 3365 3428
Department of Physics fax: 07 3365 1242
University of Queensland email: <a href="mailto:mjd_at_physics.<!--nospam-->uq.edu.au
_at_physics.uq.edu.au?Subject=Re:%20LIEF%20progress%20report%202004:%20UQ">mjd_at_physics.uq.edu.au</a>
QLD 4072, Australia http://www.physics.uq.edu.au/ap
---------- Forwarded message ----------
Background:
At the University of Queensland a cross-disciplinary team has focused
on the problem that positional uncertainty poses in terms of matching
catalogues. In 2003 (funded by an internal UQ grant) we were able to show
that machine learning techniques could be used to match the SuperCOSMOS
and HIPASS catalogues using a model we trained on a linked subset.
Personnel:
Dr. M. Drinkwater (Physics UQ)
Prof T. Downs, Dr. M. Gallagner (ITEE, UQ)
Mr. D. Rohde (Physics and ITEE, UQ)
Activities 2004:
1. We have developed a prototype system that can resolve the ambiguous
matching problem for arbitrary catalogues using a user-provided training
set. The prototype service has been installed on a 16-node
high-performance computing cluster located at UQ. [The batch server
component is now being installed; we will make this available to the IVO
group by March 2005.]
2. David Rohde, who built the prototype system, attended The US National
Virtual Observatory Summer School (Sep 2004) in order for the team to gain
greater understanding of the world context of Virtual Observatory planning
and software. His team won a competition held during the Summer School to
build a VO demonstration tool.
3. The development work has resulted in these publications/presentations:
Refereed:
`Machine Learning for Matching Catalogues', Rohde D, Drinkwater M,
Gallagher M, Downs T, Doyle M, 2004, Proceedings of Intelligent Data
Engineering and Automated Learning, 1.3177, 702-707
`Applying Machine Learning to Catalogue Matching in Astrophysics', Rohde
D, Drinkwater M, Gallagher M, Downs T, Doyle M, 2005, MNRAS, submitted
Conferences:
Intelligent Data Engineering and Automated Learning Conference 2004
`Machine Learning for Matching Catalogues'
Annual Scientific Meeting of the Astronomical Society of Australia 2004
'Machine Learning for Matching Catalogues'
###########################################################################
Received on 2005-02-02 09:34:02