Abstract-In this paper, we address dynamic clustering in high dimensional data or feature spaces as an optimization problem where multi-dimensional particle swarm optimization (MD PSO) is used to find out the true number of clusters, while fractional global best formation (FGBF) is applied to avoid local optima. Based on these techniques we then present a novel and personalized long-term ECG classification system, which addresses the problem of labeling the beats within a long-term ECG signal, known as Holter register, recorded from an individual patient. Due to the massive amount of ECG beats in a Holter register, visual inspection is quite difficult and cumbersome, if not impossible. Therefore the proposed system helps professionals to quickly and accurately diagnose any latent heart disease by examining only the representative beats (the so called master key-beats) each of which is representing a cluster of homogeneous (similar) beats. We tested the system on a benchmark database where the beats of each Holter register have been manually labeled by cardiologists. The selection of the right master key-beats is the key factor for achieving a highly accurate classification and the proposed systematic approach produced results that were consistent with the manual labels with 99.5% average accuracy, which basically shows the efficiency of the system.
I. INTRODUCTIONATA clustering is a multi-modal problem especially in high dimensions. There are many suboptimal solutions and well-known deterministic methods such as K-means,, etc. are susceptible to get trapped to the closest local optimum since they are nothing but greedy descent methods, which start from a random point in the solution space and perform a localized search. Furthermore, all the mentioned clustering algorithms require the number of clusters to be specified in advance.Multi-dimensional particle swarm optimization (MD PSO) [6] is a dynamic extension of the basic particle swarm optimization (bPSO) algorithm, where the particles can make inter-dimensional passes during their search. Thus they are not restricted to search for an optimal solution in a single dimension, but the optimal solution dimension is solved in parallel. As a stochastic optimization technique MD PSO can avoid some local optima but is still susceptible 1 This work was supported by the Academy of Finland, project No. 213462 (Finnish Centre of Excellence Program (2006 -2011 to premature convergence due to lack of divergence. Fractional global best formation (FGBF), on the other hand, is an effective cure for the premature convergence. It basically collects all promising components of the particle positions and fractionally creates an artificial global best solution that has a potential to be a better guide than the best solution found by the swarm, which is used in the bPSO. Combined, MD PSO and FGBF form an efficient dynamic clustering technique with a high resistance to local optima.Long-term continuous electrocardiogram (ECG) monitoring and recording, also known as Holter electrocard...