Using Fuzzy Clustering with Bioinformatics Data

The article describes a research about fuzzy clustering algorithms, their creation and classification with the goal to determine the possibilities to use them in bioinformatics data clustering to find the membership of each record to a class. The study uses sixteen data sets used in previous studies by the authors and other researchers. Experiments were carried out using fuzzy c-means clustering method. The first section of the article gives an overview of the historical development of fuzzy clustering algorithms, their classification as well as the hypothesis that fuzzy clustering algorithms can be used to construct membership functions. The second section gives the description of the applied algorithm and the sixteen data sets used in the experiments. The third section gives a summary of the performed experiments and their results. And finally conclusions are drawn about the use of the algorithms in the clustering of bioinformatics data. The fourth section gives the overall conclusions and describes the further research directions. It is proven that fuzzy clustering algorithms (including the most popular – fuzzy c-means) can be used in membership function construction. Therefore fuzzy c-means algorithm with slight modifications can be used to construct membership functions of separate record attributes