In this paper, we extend the conventional vector quantization by incorporating a vigilance parameter, which steers the tradeoff between plasticity and stability during incremental...
Data warehouses provide a great deal of opportunities for performing data mining tasks such as classification and clustering. Typically, updates are collected and applied to the d...
Relational clustering has attracted more and more attention due to its phenomenal impact in various important applications which involve multi-type interrelated data objects, such...
Background: The recent advancement of microarray technology with lower noise and better affordability makes it possible to determine expression of several thousand genes simultane...
Raja Loganantharaj, Satish Cheepala, John Clifford
If there are more clusters than the ideal, each intrinsic cluster will be split into several subsets. Theoretically, this split can be arbitrary and neighboring data points have a ...