Use OPTICS as it can find irregularily shaped clusters - Maybe we can get better cluster validation indices with that. Especially interesting is the performance given the large cluster overlap in the latent profile analysis.
An important first step is the definition of a grid with values foe the epsilon and minpoints parameters.
Use OPTICS as it can find irregularily shaped clusters - Maybe we can get better cluster validation indices with that. Especially interesting is the performance given the large cluster overlap in the latent profile analysis. An important first step is the definition of a grid with values foe the epsilon and minpoints parameters.