Loading ...
Sorry, an error occurred while loading the content.

[Fwd: BISC: Four research positions at the European Centre for Soft Computing]

Expand Messages
  • Tina Yu
    ... Subject: BISC: Four research positions at the European Centre for Soft Computing From: Masoud Nikravesh Date: Wed, May
    Message 1 of 1 , May 31, 2006
    • 0 Attachment
      ---------------------------- Original Message ----------------------------
      Subject: BISC: Four research positions at the European Centre for Soft
      Computing From: "Masoud Nikravesh" <Nikravesh@...>
      Date: Wed, May 31, 2006 2:33 pm
      To: "BISC Program-UCB" <BISC-Group@...>
      --------------------------------------------------------------------------

      *********************************************************************
      Berkeley Initiative in Soft Computing (BISC)
      *********************************************************************

      Four research positions are available at the European Centre for Soft
      Computing (ECSC), recently created in Mieres, Asturias, Spain
      (http://www.softcomputing.es <http://www.softcomputing.es/>). Two of these
      vacancies correspond to the Intelligent Data Analysis (IDA)
      Research Unit for Associate and/or Postdoctoral researchers, and the other
      two are for Postdoctoral researchers in the Applications of Fuzzy Logic
      and Evolutionary Algorithms (AFE) Research Unit.

      Applicants should send (before June 30, 2006) an e-mail to
      positions@... <mailto:positions@...>
      identifying the position/s to which they apply and including their
      curriculum vitae. Final results will be known by the end of July 2006.

      Apart from the specific positions’ requirements, applicants should have
      fluent English skills (written and spoken).

      For additional information on a particular position use the
      corresponding contact information.

      AVAILABLE POSITIONS

      POSITION 1

      RESEARCH UNIT: Intelligent Data Analysis
      PRINCIPAL RESEARCHER: Christian Borgelt
      (christian.borgelt@...) POSITION: Associate or Postdoctoral
      researcher
      TOPIC: Molecular Fragment Mining

      DESCRIPTION: A frequent task in biochemistry is the search for common
      features in large sets of molecules. Examples are drug discovery, where
      one is interested in identifying properties shared by molecules that have
      been classified as "active" (and rarely shared by those classified as
      "inactive") w.r.t., for example, the protection of human cells
      against a virus, and compound synthesis, where one tries to identify
      properties that enable or inhibit the synthesis of new molecules, so that
      one can predict the chances for a successful synthesis.

      Earlier research carried out by the principal researcher of the
      Intelligent Data Analysis Unit focused on an approach that models
      molecules as attributed graphs, thus taking the connection structure,
      though not the 3-dimensional structure into account. The resulting set of
      graphs is then searched for common subgraphs, that is, molecular fragments
      that appear with a user-specified minimum frequency. The result of this
      research is the MoSS/MoFa algorithm, which has been implemented in Java.

      Future research in this direction plans to take additional features than
      just the connection structure into account and thus to make the search for
      interesting substructures more flexible, but also more geared to the needs
      of (bio)chemists. Such features include the 3D-structure of
      molecules, charge distribution over molecules, properties like
      solubility in water etc. Especially interesting is the possibility

      to replace certain atoms by certain others or even groups of atoms (at
      least in certain contexts) without changing the chemical properties too
      much. This could lead to approximate or "fuzzy" fragments, which could be
      more meaningful to the (bio)chemist or drug designer.

      REQUIREMENTS:

      Applicants should
      - have a strong background in computational chemistry, biochemistry or
      pharmacy
      - be well-educated in statistical and computational methods
      - have at least some basic knowledge of graph theory and algorithms

      Holding a Ph.D. in (bio-)chemistry or a closely related subject is highly
      appreciated.
      Programming skills, especially in Java, would be nice, but are not
      required, as the
      focus is more on method development and algorithm design than on
      implementation.

      ------------------------------------------------------------------------

      POSITION 2

      RESEARCH UNIT: Intelligent Data Analysis
      PRINCIPAL RESEARCHER: Christian Borgelt
      (christian.borgelt@...) POSITION: Associate or Postdoctoral
      researcher
      TOPIC: Diagnosis of Technical Systems with Graphical Models

      DESCRIPTION: Although graphical models (like Bayes networks or Markov
      networks) have a long tradition of being used for diagnosis tasks, a
      proper methodology for constructing a graphical model more or less
      automatically from a technical description of the system is still
      missing. The vision is to have a software system, which, if fed with a
      description of the mechanical and electrical components of the system and
      their interaction produces a graphical model that can be used to find
      faulty parts from observations and measurements carried out on the system.

      In earlier research carried out by the principal researcher of the
      Intelligent Data Analysis Unit some initial ideas and methods to
      automatically construct a Markov network from a description of an analog
      electrical circuit have been developed. An important focus of this
      research was to identify so-called "soft faults", that is, the deviation
      of properties of electrical elements from their nominal values rather than
      their complete failure.

      Future research in this direction holds several challenges in stock: One
      needs to develop an automatic and efficient procedure to produce a "good"
      division of the circuit (or generally, a technical system) so that
      diagnosis with the resulting graphical model is computationally feasible.
      Better ways of handling deterministic relationships between variables need
      to be developed (the state of the art can only handle piecewise linear
      relationships). If knowledge about relationships or measurement results
      can be obtained only with a certain degree of
      imprecision, an extension to possibilistic graphical models can be tried etc.

      REQUIREMENTS:

      Applicants should:
      - have a strong background in probability theory and especially the theory
      of graphical models (Bayes networks and Markov networks)
      - not be averse to considering alternatives like possibilistic networks. -
      have at least basic knowledge of physical and electrical laws as well as
      of core approaches to diagnosis

      It would be highly appreciated if the applicant has already worked on a
      technical diagnosis problem, preferably using graphical models.
      Programming skills in Java or C are recommended as a strong emphasis will
      be on implementing the developed methods.

      ------------------------------------------------------------------------

      POSITION 3

      RESEARCH UNIT: Applications of Fuzzy Logic and Evolutionary Algorithms
      PRINCIPAL RESEARCHER: Oscar Cordón (oscar.cordon@...)
      POSITION: Postdoctoral researcher
      TOPIC: Automatic forensic identification based on soft computing and
      computer vision

      REQUIREMENTS:

      Applicants should:
      - hold a PhD on Computer Science, Electrical Engineering, Math., or
      related disciplines,
      - have a track record of research in soft computing techniques,
      especially in evolutionary computation,
      - have a good background on computer vision, mainly on image processing,
      image registration and 3D model building,
      - have basic computer programming skills.

      JOB PROFILE AND ECONOMIC CONDITIONS:

      The position is to be filled in under the form of a three year work
      contract, starting by September/October 2006.
      Annual 32,000 euros gross salary.
      Full Social Security rights, including health assistance for the
      researcher and her/his family.

      ------------------------------------------------------------------------

      POSITION 4

      RESEARCH UNIT: Applications of Fuzzy Logic and Evolutionary Algorithms
      PRINCIPAL RESEARCHER: Oscar Cordón (oscar.cordon@...)
      POSITION: Postdoctoral researcher
      TOPIC: Applications of fuzzy logic/fuzzy systems and evolutionary
      algorithms to:
      - visual science map design and mining
      - hierarchical fuzzy models design within the interpretability-accuracy
      trade-off framework

      REQUIREMENTS:

      Applicants should:
      - hold a PhD on Computer Science, Electrical Engineering, Math., or
      related disciplines,
      - have a track record of research in fuzzy systems and evolutionary
      algorithms from the machine learning viewpoint,
      - have basic computer programming skills.

      Some basic knowledge of graph theory and algorithm theory would be nice,
      but is not mandatory.

      JOB PROFILE AND ECONOMIC CONDITIONS:

      The position is to be filled in under the form of a three year work
      contract, starting by September/October 2006.
      Annual 32,000 euros gross salary.
      Full Social Security rights, including health assistance for the
      researcher and her/his family.



      -------------------------------------------------------------------- If
      you ever want to remove yourself from this mailing list,
      you can send mail to <Majordomo@...> with the following
      command in the body of your email message:
      unsubscribe bisc-group
      or from another account,
      unsubscribe bisc-group <your_email_adress>



      --
      Dr. Tina Yu
      Associate Professor
      http://www.cs.mun.ca/~tinayu/index.htm
      Tel: 709-737-6943
      Fax: 709-737-2009
      Department of Computer Science
      Memorial University of Newfoundland
      St. John's, NL A1B 3X5
      Canada
    Your message has been successfully submitted and would be delivered to recipients shortly.