Onesfole Kurama


Current Organisation

Research Interest
Kurama Onesfole
New fuzzy PCA algorithms and similarity classifier development in data anal-
ysis .
Lappeenranta, 2014.
In this research we contribute to the development of a new classifier, which is quite robust for the amount of data and performs very well with small amounts of data. This is done by concentrating on developing arobust PCA algorithms from which we develop new methods. The other aspect is to fit the processing methods to the similarity classifier as well as possible development of the classifier. In our previous article, "Similarity classifier with ordered weighted averaging operators", the similarity classifier was extended to cover also the ordered averaging operators(OWA) as opposed to the well known generalized mean approach. The proposed method was applied to the real world medical data sets and the performance was very good compared to those
got earlier using the generalized mean approach. In our next work, the concentration is put on another aggregation by use of the lambda-averaging operator. In the broad sense, classification is said to be the 'heart' of any pattern recognition system and is an important part in many machine vision systems. In medical diagnostic and epidemiological research, classification has a very important role, for example, diseases must be recognized before treatment can be started, in such systems accuracy of the classifier has a crucial role since a small improvement in accuracy can save human lives. Systems require constant monitoring, like in quality control models, classifiers are in great demand since faulty products must be recognized from the assemble line. Other major roles of Classifiers like on-line classification will be tackled in this research.
Key words: Similarity classifier, OWA operators, Quantifier guided aggregation, classification, medical data, Principal Component Analysis(PCA).

Teaching Experience
I have taught as an Assistant Lecturer at the Department of Mathematics-Makerere University-Uganda for 3 years.

Publications
A similarity classifier with generalized ordered weighted averaging operator (2017)
A4 Conference proceedings
Kurama Onesfole, Luukka Pasi, Collan Mikael
Proceedings of the 17th International Fuzzy Systems Association world congress (IFSA-SCIS), 2017
An n-ary λ-averaging based similarity classifier (2016)
A1 Journal article (refereed), original research
Kurama Onesfole, Luukka Pasi, Collan Mikael
International Journal of Applied Mathematics and Computer Science
On a weighted ordered weighted averaging based similarity classifier (2016)
A4 Conference proceedings
Kurama Onesfole, Luukka Pasi, Collan Mikael
Proceedings of the IEEE 17th International Symposium on Computational Intelligence and Informatics (CINTI 2016)
Similarity classifier with weighted ordered weighted averaging operator (2016)
A1 Journal article (refereed), original research
Kurama Onesfole, Luukka Pasi, Collan Mikael
Fuzzy Economic Review




Last updated on 2017-22-03 at 13:30