Seminar on Computer Science and Applied Mathematics
Upravni odbor Matematickog instituta SANU je na nedavnoj sednici doneo odluku da se dosadasnji Seminar za primenjenu matematiku, sada nazove Seminar za racunarstvo i primenjenu matematiku, a u cilju potenciranja znacaja racunarstva kao jedne od oblasti delatnosti Instituta. Istovremeno, Upravni odbor doneo je odluku o osnivanju Odeljenja za racunarstvo i primenjenu matematiku i vezao rad novog odeljenja za rad Seminara za racunarstvo i primenjenu matematiku.
Ponedeljak, 22.07.2013. u 12:00, Sala 301f, MI SANU:
prof. Zoran Obradovic, Director, Data Analytics and Biomedical Informatics
Professor, Computer and Information Sciences Department,
Professor, Statistics Department, Fox School of Business (secondary appointment),
CONTINUOUS CONDITIONAL RANDOM FIELDS FOR EFFICIENT REGRESSION IN LARGE FULLY CONNECTED GRAPHS
When used for structured regression, powerful Conditional Random Fields
(CRFs) are typically restricted to modeling effects of interactions among
examples in local neighborhoods. Using more expressive representation would
result in dense graphs, making these methods impractical for large-scale
applications. To address this issue, we propose an effective CRF model with
linear scale-up properties regarding approximate learning and inference for
structured regression on large, fully connected graphs. The proposed method
is validated on real-world large-scale problems of image denoising and
remote sensing. In conducted experiments, we demonstrated that dense
connectivity provides an improvement in prediction accuracy. Inference time
of less than ten seconds on graphs with millions of nodes and trillions of
edges makes the proposed model an attractive tool for large-scale,
structured regression problems.
This is joint research with Vladan Radosavljevic, Kosta Ristovski, and Slobodan Vucetic and the results will be published at Proc. Twenty-Seventh AAAI Conference on Artificial Intelligence (AAAI-13), Bellevue, Washington, July 2013.
Utorak, 23.07.2013. u 14:15h, Sala 301f, MI SANU:
Snezana Minic, Simon Fraser University, Vancouver, Canada
IMAGE ACQUISITION SCHEDULING AND CLOUD AVOIDANCE FOR AGILE HI-RESOLUTION EARTH OBSERVING OPTICAL SATELLITE
Abstract: An agile satellite that can rapidly slew its imager boresight in
both roll and pitch angles can be used for obtaining cloud-free optical
images. After detecting the cloud coverage for the area over which the
satellite will pass in the next few minutes, the image acquisition
scheduling problem is solved on-board the satellite. We present a
mathematical programming formulation of the problem and an experimental
study comparing heuristic solution approaches. The problem instances were
generated with varying cloud coverage scenarios simulating the weather
conditions over a real geographic area.
Joint work with Joe Steyn, MDA Systems.