| Title |
Daugiamačių duomenų rinkinių tyrimas taikant modifikuotą Renyi-CLIPx klasterizacijos algoritmą |
| Translation of Title |
Multidimensional data analysis using modified Renyi-CLIPx conceptual clustering algorithm. |
| Authors |
Narščius, Aleksas |
| Full Text |
|
| Pages |
76 |
| Keywords [eng] |
Clustering ; conceptual clustering ; clustering rules ; virtual learning system ; modified Renyi-CLIPx algorithm |
| Abstract [eng] |
In this master’s paper a great attention is paid for multidimensional data analysis by conceptual clustering algorithms. This analysis includes two aspects: discovering of clusters and clustering rules form various data sets. These rules help to decide to which cluster newly appearing object has to be assigned. We also are interested in analysis of virtual learning system’s data sets which contains such information as grades of students. Clustering results are tested by using training data which real clusters are well known. Made results indicates about a very good clustering solution. Discovered algorithms and results of experiments are important for developing of virtual learning system. We suppose that it is useful for making better self-studies with modern communication technologies. Modified Renyi-CLIPx clustering algorithm is proposed as result of analysis Renyi entropy, ITERATE, CLIP3 and CLIP4 clustering algorithms. It is used to discover better clusters and create rules for virtual learning and other data sets. |
| Type |
Master thesis |
| Language |
Lithuanian |
| Publication date |
2008 |
