BEGIN:VCALENDAR
VERSION:2.0
METHOD:PUBLISH
X-WR-CALNAME: Calendar | Seminars
PRODID:-//Drupal iCal API//EN
BEGIN:VEVENT
UID:calendar.384.field_seminar_date.0.0
SUMMARY:Sparse Representation based on Dictionary Learning
DTSTAMP:20191014T173011Z
DTSTART:20191016T091500Z
DTEND:20191016T091500Z
URL;VALUE=URI:http://seminars.nersc.no/seminar/sparse-representation-based-dictionary-learning
LOCATION:Lecture room\, Ground Floor\, NERSC
DESCRIPTION:This work presents the use of the Dictionary Learning method for a sparse \n
representation of 4D seismic data. We consider a trade-off between the number \n
of nonzero coefficients retained in the sparse data representation\, the \n
computational cost\, and how well we can capture the main features of the \n
original 4D seismic signal. K-SVD is an iterative algorithm used in \n
Dictionary Learning that alternates between the calculation of the sparse \n
representation vector and dictionary update. The algorithm starts with the \n
definition of an initial dictionary (Discrete Cosine Transform\, for \n
instance). To calculate the sparse representation vector\, one can use the \n
Orthogonal Matching Pursuit (OMP) algorithm and constrain the problem into \n
two distinct approaches: (1) sparsity-constrained\; and (2) error-constrained. \n
Finally\, it is possible to update the dictionary through SVD. We evaluated \n
the influence of critical parameters of the algorithm (dictionary size\, \n
number of iterations\, patch size\, and training dataset size). Results showed \n
that the dictionary learning method can capture the main features of the \n
original 4D seismic signal with the sparse representation. However\, the \n
number of nonzero coefficients retained is highly dependent on the selected \n
parameters. Therefore\, they need to be carefully determined to obtain a \n
reasonable amount of nonzero coefficients retained.
END:VEVENT
BEGIN:VEVENT
UID:calendar.386.field_seminar_date.0.1
SUMMARY:Chasing Water: How ocean currents transport plastic and plankton around the \n
globe
DTSTAMP:20191014T173011Z
DTSTART:20191021T091500Z
DTEND:20191021T091500Z
URL;VALUE=URI:http://seminars.nersc.no/seminar/chasing-water-how-ocean-currents-transport-plastic-and-plankton-around-globe
LOCATION:Lecture room\, Ground Floor\, NERSC
DESCRIPTION:The ocean is in constant motion\, with water circulating within and flowing \n
between basins. As the water moves around\, it caries heat and nutrients\, as \n
well as planktonic organisms and plastic litter around the globe.\n
\n
The most natural way to study the pathways of water and the connections \n
between ocean basins is using particle trajectories. The trajectories can \n
come from computing of virtual floats in high-resolution ocean models.\n
\n
In this seminar\, I'll give an overview of some recent work with Lagrangian \n
particles. I will introduce our new open-source oceanparcels.org framework. I \n
will show applications to marine microbiology and ecology\, palaeoclimatology \n
and plastic pollution. Central to each of these studies is the question on \n
how connected the different ocean basins are\, and on what time scales water \n
flows between the different regions of the ocean.
END:VEVENT
END:VCALENDAR