ACM Computing Surveys (CSUR)
Knowledge discovery by application of rough set models
Rough set methods and applications
Rough Sets: Theoretical Aspects of Reasoning about Data
Rough Sets: Theoretical Aspects of Reasoning about Data
Pattern Recognition with Fuzzy Objective Function Algorithms
Pattern Recognition with Fuzzy Objective Function Algorithms
On Clustering Validation Techniques
Journal of Intelligent Information Systems
Rough Set Data Mining of Diabetes Data
ISMIS '99 Proceedings of the 11th International Symposium on Foundations of Intelligent Systems
Knowledge-Based Clustering: From Data to Information Granules
Knowledge-Based Clustering: From Data to Information Granules
Data Mining: Practical Machine Learning Tools and Techniques, Second Edition (Morgan Kaufmann Series in Data Management Systems)
Modeling of High Quality Granules
RSEISP '07 Proceedings of the international conference on Rough Sets and Intelligent Systems Paradigms
Fast Discovery of Minimal Sets of Attributes Functionally Determining a Decision Attribute
RSEISP '07 Proceedings of the international conference on Rough Sets and Intelligent Systems Paradigms
Rough Granular Computing in Knowledge Discovery and Data Mining
Rough Granular Computing in Knowledge Discovery and Data Mining
Hi-index | 0.00 |
Granular computing is a multidisciplinary theory rapidly developed in recent years. It provides a conceptual framework for many research fields, among others data mining. Data mining techniques and algorithms focus on knowledge discovery from data. When data labels are unknown one can use methods of exploratory data analysis called clustering algorithms. Clustering algorithms are also useful to find hidden dependencies and patterns in data. In this article granular computing and clustering were implemented in information granulation system SOSIG and applied to exploration of real medical data set. Data granulation in the system can be performed on different levels of resolution. Thereby the granules composed of clusters reflect relationship between objects on distinct levels of details. The clustering in SOSIG is generated automatically - there is no requirement to give a number of groups for division. It eliminates problems present in popular clustering algorithms like selection of correct number of clusters and evaluation of created partitioning. The difficulties are encountered in most partitioning as well as hierarchical methods reducing their practical application. Additionally, this article contains solution generated by SOSIG in comparison with clustering results of algorithms: k-means, hierarchical, EM and DBSCAN. There are used quality indices such as Dunn's, DB, CDbw and SI.