POSTDOCTORAL RESEARCH OPPORTUNITY - BIOMEDICAL DATA ANALYSIS

[Pizzi, Nick]

POSTDOCTORAL RESEARCH OPPORTUNITY - BIOMEDICAL DATA ANALYSIS
GRANULAR COMPUTING AND SPARSE CODING FOR THE MODELING, ANALYSIS, AND
CLASSIFICATION OF HIGH-DIMENSIONAL BIOMEDICAL DATA

Pending final approval, a postdoctoral position is available at the
University of Manitoba to investigate sparse coding strategies for the
modeling, analysis, and classification of high-dimensional biomedical
data. This is an NSERC-funded strategic research project directed by
Prof. W. Pedrycz (U. Alberta), Dr. N. Pizzi (National Research Council &
U. Manitoba), and Dr. M. Alexander (National Research Council). The
position, which will focus on sparse coding analysis, is for a one-year
term with a possible one year renewal.

Biomedical data classification demands a reduction of the original
data's dimensionality without sacrificing domain-specific significance.
Sparse representation (SR) addresses the former issue and granular
computing the latter. The fundamental tasks of feature extraction,
classification, and noise/artefact suppression all benefit by being
carried out in a sparse domain.

Since wavelets have become a powerful computational tool in signal
processing, there has been a growing interest in the SR of data. The
transformations that accomplish optimal sparseness will depend on the
data itself, and may, for example, be defined by optimizing its entropy
over a finitely- (or infinitely) over-complete "dictionary" of basis
functions. The computational task of finding the optimal SR for given
data is accomplished in two stages: (i) Defining an over-complete
dictionary by constructing a sufficiently broad class of basis
functions; (ii) Projecting the data onto a particular subset of the
over-complete dictionary that optimizes a pre-defined measure of
sparseness. Computationally efficient algorithms need to be considered
for (i), and the optimization procedure for (ii) may in some cases
involve substantial computation. Both (i) and (ii) are topics of ongoing
research.

Required background:
-	Recent Ph.D. graduate
-	Training and/or research experience in signal processing
-	Background and/or experience in wavelets and their application
to data analysis
-	Training and/or research experience in statistics, including
practical knowledge of analysis of large datasets
-	Proficiency in C/C++ and experience with using MatLab/IDL

Please send a curriculum vitae, expression of interest (including
earliest start date), and the names and e-mail addresses of two
references to Dr. Pizzi at pizzi at nrc.ca. Your curriculum vitae should
include a list of recent publications. Please outline your interest in
this project, how it is related to work that you have done, and what
special expertise you would bring to the project.

-------------------------------------------------------------------
Nicolino Pizzi, PhD                                    pizzi at nrc.ca
Senior Research Officer
Institute for Biodiagnostics
National Research Council Canada
435 Ellice Avenue                               Ph: +1 204 983 8842
Winnipeg MB, R3B 1Y6, Canada                    Fx: +1 204 984 5472

Adjunct Professor
Computer Science Department
& Electrical and Computer
  Engineering Department
University of Manitoba            http://www.ee.umanitoba.ca/~pizzi
-------------------------------------------------------------------