Skip to content

Ž. Meškauskas “Expert risk evaluation system based on explainable artificial intelligence” doctoral dissertation defense

Thesis defense

Author, Institution: Žygimantas Meškauskas, Kaunas University of Technology

Science area, field of science: Technological Sciences, Informatics Engineering, T007

Scientific Supervisor: Prof. Dr. Egidijus Kazanavičius (Kaunas University of Technology, Technological Sciences, Informatics Engineering, T007)

Dissertation Defence Board of Informatics Engineering Science Field:
Prof. Dr. Hab. Rimvydas Simutis (Kaunas University of Technology, Technological Sciences, Informatics Engineering, T007) – chairperson
Prof. Dr. Hab. Gintautas Dzemyda  (Vilnius University, Technological Sciences, Informatics Engineering, T007)
Assoc. Prof. Dr. Nikolaj Goranin  (Vilnius Gediminas Technical University, Technological Sciences, Informatics Engineering, T007)
Prof. Dr. Hab. Artūras Kaklauskas (Vilnius Gediminas Technical University, Technological Sciences, Informatics Engineering, T007)
Dr. Zbigniew Marszałek (Silesian University of Technology, Poland, Natural Sciences, Informatics, N009)

 

Dissertation defence meeting will be at M7 Hall at The Campus Library of Kaunas University of Technology (Studentų 48–M7, Kaunas)

 

The doctoral dissertation is available at the library of Kaunas University of Technology (K. Donelaičio g. 20, Kaunas)

 

Annotation: This work proposes the possibility of extending systems based on expert estimates for more efficient knowledge gathering and presentation, and for applicability to risk assessment. One of the most widely used methods for situation assessment is the SWOT analysis of strengths, weaknesses, opportunities, and threats. This method uses the human experience of experts to identify the weaknesses and strengths of the situation under consideration. Classically, the situation description using the SWOT analysis method requires experts to provide quantitative estimates, which is often a complex, imprecise, and slow process, so an addition to the classical method has been proposed by adding a “computing with words” paradigm and allowing experts to provide estimates in natural language words from a defined vocabulary with a level of uncertainty. The proposed extension of the classical approach introduces elements of soft computing and word processing by applying fuzzy logic to work with imprecise data. By analyzing the scope of the proposed addition, the work reviews existing risk assessment concepts, methods, frameworks and develops a new concept of risk, which is the basis for the proposed domain-independent new risk assessment method. The proposed risk assessment method has been applied in the developed systems and its effectiveness has been assessed.

June 27 d. 09:00

M7 Hall at The Campus Library of Kaunas University of Technology (Studentų 48–M7, Kaunas)

Įtraukti į iCal
Suggest an Event