1. What are the contributions in "A framework for initialising a dynamic clustering algorithm: art2-a" ?
In this paper the authors introduce an adaptation of the ART2-A method within a separation and concordance ( SeCo ) framework which has been shown to identify stable and reproducible solutions from repeated initialisations that also provides evidence for an appropriate number of initial clusters that best calibrates the algorithm with the data presented.
read more
2. What future works have the authors mentioned in the paper "A framework for initialising a dynamic clustering algorithm: art2-a" ?
Using a dual measure approach to initialisation of this algorithm provides a good starting position to using the algorithm in a big data environment, allowing for robust partitions to be used to cluster incoming data whilst still allowing for the possibility of new cohorts being identified over time and in such a context ART2A provides a realistic alternative to hierarchical, k-means or other similar Expectation-Maximisation like algorithms.. Identifying data which does not fit existing prototypes is not enough however, and whilst this work provides a good starting point for clustering of big data in healthcare, there is a need to ensure that such new prototypes are in fact signal and not noise, and further work is required to evaluate the performance of the method in actual big data scenarios and benchmark the effect of introducing new cohorts into a previously trained structure.. A further step might be to include a hierarchical application of this approach to explore the existence of substructure within the membership of these groups.
read more
