en:konkurs:laureaci2013

Differences

This shows you the differences between two versions of the page.


en:konkurs:laureaci2013 [2020/06/17 21:43] (current) – created - external edit 127.0.0.1
Line 1: Line 1:
 +~~NOTOC~~
  
 +====== Winners of the PSSI Award for the Best Ph. D. Dissertation in Artificial Intelligence in 2013 ======
 +
 +{{  :pl:konkurs:kajdanowicz.jpg|Tomasz Kajdanowicz}}
 +
 +===== 2013 Polish Artificial Intelligence Society Award for the Best Ph. D. Dissertation in Artificial Intelligence =====
 +
 +**Dr. Tomasz Kajdanowicz**, Wrocław University of Technology \\
 +"//[[http://www.kajdanowicz.com/library/pssi2012konkurs-Kajdanowicz-rozprawa.pdf|Classification Methods based on Multi-label Problem Transformation]]//" \\
 +Supervisor: **Prof. Przemysław Kazienko.**
 +
 +This dissertation concerns a problem of multi-label classification. Multi-label classification allows mapping of multiple class labels assigned at the same time to a single data instance. In order to face common problems in multi-label classification based on problem transformation, several new solutions providing more accurate classification in reasonable time are proposed in the dissertation:  Classifier Chain, AdaBoostSeq and learning framework based on Error Correcting Output Codes. In the dissertation a complimentary assessment of proposed methods as well as discussion on their efficiency and computational complexity in comparison to state-of-the-art methods are provided.
 +
 +\\
 +
 +\\
 +
 +{{  :pl:konkurs:bigaj.jpg|Piotr Bigaj}}
 +
 +===== Honourable mention in the competition for the 2013 Polish Artificial Intelligence Society Award for the Best Ph. D. Dissertation in Artificial Intelligence ===== 
 +**Dr. Piotr Bigaj**, Systems Research Institute, Polish Academy of Sciences \\
 +"//[[https://www.dropbox.com/s/rxbc8eg0jtesvfm/Doktorat_wersja_finalna_bez_podzienkowan.pdf|A memetic algorithm for the global path planning with movement constraints for a non-holonomic mobile robot]]//" \\
 +Supervisor: **Prof. Janusz Kacprzyk**
 +
 +The paper presents an innovative solution to one of the most important problems of mobile robotics which is global path planning (Global Path Planning GPP). This issue concerns solving the optimization problem with constraints. Constraints are associated with a set of rules for path planning including traffic restrictions imposed by non-holonomicity of the robot, while the optimization aspect is reflected in the fact that the search focuses on the shortest, most "effective" path. The thesis that was set and proved during this research  was that biologically inspired algorithms, in particular, memetic algorithms, can be an effective tool for global path planning of robots, both for holonomic and non-holonomic mobile robots.
 +
 +\\
 +
 +\\
 +
 +
 +{{  :pl:konkurs:lech.jpg|Michał Lech}}
 +
 +===== Honourable mention in the competition for the 2013 Polish Artificial Intelligence Society Award for the Best Ph. D. Dissertation in Artificial Intelligence ===== 
 +**Dr. Michał Lech**, Gdańsk University of Technology \\
 +"//[[http://sound.eti.pg.gda.pl/~mlech/pssi2013konkurs_lech_rozprawa.pdf|A method and algorithms for controlling sound mixing processes by hand gestures recognized in a video stream]]//" \\
 +Supervisor: **Prof. Bożena Kostek**
 +
 +In the scope of the dissertation the novel sound mixing system enabling a user to mix via hand gestures recognized in the varying image displayed by the multimedia projector has been developed. The gesture recognition model has been based on comparative analysis of the video stream displayed by the projector and video stream captured from the camera. The model employs fuzzy logic, support vector machines and author’s image processing method, which ensures high efficacy of gesture recognition without using devices such as infrared sensors and emitters.
 +
 +\\
 +
 +\\
 +
 +
 +{{  :pl:konkurs:napierala.jpg|Krystyna Napierała}}
 +
 +===== Honourable mention in the competition for the 2013 Polish Artificial Intelligence Society Award for the Best Ph. D. Dissertation in Artificial Intelligence ===== 
 +**Dr. Krystyna Napierała**, Poznań University of Technology  \\
 +"//[[http://www.cs.put.poznan.pl/knapierala/misc/pssi2013konkursNapierala-rozprawa.pdf|Improving Rule Classifiers For Imbalanced Data]]//" \\
 +Supervisor: **Prof. Jerzy Stefanowski**
 +
 +The dissertation concerns rule learning from imbalanced data, in which one of the classes contains much fewer examples than the other classes. Class imbalance, often observed in medicine, banking and technical diagnostics, is difficult for most learning methods. The first part of the thesis analyses the difficulty factors influencing the learning abilities. Then, two new rule learning algorithms are proposed: (1) taking into account the local data distributions and (2) incorporating expert's explanations for some learning examples.
 +
 +\\
 +
 +\\