Ph.D. grant available - Project Rank!

Carlos Soares csoares at fep.up.pt
Do Aug 21 19:30:19 CEST 2008


_*UNIVERSIDADE DO PORTO*_


  RESEARCH GRANT

The Rank! Project (http://www.liaad.up.pt/csoares/rank) is looking for 
researchers in Artificial Intelligence/Data Mining to do work in Machine 
Learning methods for predicting rankings.


    Description

The Rank! Project (PTDC/EIA/81178/2006) is funded by the Portuguese 
Government (Fundação para a Ciência e Tecnologia). The main goal is to 
develop of a methodology for predicting rankings.

The problem of learning to predict rankings is similar to the problem of 
supervised classification in Machine Learning/Data Mining. However, the 
goal is not to predict which class each example belong to but to rank 
the classes for each example. There are several applications that 
involve predicting rankings (e.g., recommender systems on the web, stock 
trading, medical diagnosis, marketing and sales), thus, the amount of 
research dedicated to this problem is increasing.

The grantees will work on one or more of the tasks involved in 
developing a methodology for predicting rankings, namely:

    *

      formalization of the problem

    *

      development/adaptation of Machine Learning methods for the problem
      of predicting rankings

    *

      development of evaluation measures for the problem of predicting
      rankings

    *

      application of the methodology developed on practical problems

The project and the grants are managed by the Faculdade de Economia do 
Porto and the work will be carried out at LIAAD-INESC Porto L.A. 
(http://www.liaad.up.pt), in Porto, Portugal.

More information:

    *

      http://www.liaad.up.pt/csoares/rank

    *

      contact Carlos Soares (+351-223392095 or csoares at fep.up.pt)


    Profile

Eligible academic levels include undergraduate, B.Sc., Lic. and M.Sc. 
from any nationality.

Minimal requirements:

    *

      Good knowledge of Artificial Intelligence and Machine Learning

    *

      Knowledge of programming

    *

      Knowledge of Statistics

    *

      Strong motivation to work in research

    *

      Interest in pursuing studies (Ph.D.)

Desirable abilities:

    *

      Programming in R

    *

      Experience in the problem of learning to predict rankings

Selection criteria:

    *

      Analysis of curriculum vitae

    *

      Suitability of the candidate's profile for the position

    *

      Optional interview (Note: The Faculdade de Economia do Porto may
      only interview the candidates who are better positioned after
      analysis of the application documents).

The Faculdade de Economia do Porto may not hire any candidate, in case 
none of the applicants has a suitable profile.


    Conditions

The grant is predicted to start in 29 September 2008 and has a duration 
of 6 or 12 months, depending on the profile of the selected applicant, 
and may be extended up to 3 years. Full time dedication is required, 
according to Portuguese Law (Lei nº 40/2004).

The salary is defined in the corresponding FCT regulations, with current 
approximate values of (http://www.fct.mctes.pt/pt/apoios/bolsas/valores/):

    *

      385 € per month for undergraduate and B.Sc.

    *

      745 € per month for Lic.

    *

      980 € per month for M.Sc.

The funding organization will be in charge of costs with social welfare, 
according to the Portuguese Law (Lei nº 40/2004) and an insurance for 
personal injuries, according to the regulation for grants of the 
Universidade do Porto 
(http://sigarra.up.pt/rscup/noticias_geral.ver_noticia?p_nr=168).


    Application

Submit via email to D. Lídia Soares (lidia at fep.up.pt) until *15 
September 2008*, the following documents:

    *

      application letter

    *

      curriculum vitae

    *

      copy of certificate of academic level or courses completed in
      current level

*This project is funded by FCT, through PIDDAC.*


Faculty of Economics University of Porto


Dean of the Faculty of Economics



(Prof. Doutor José da Silva Costa)






Mehr Informationen über die Mailingliste IFI-CI-Event