Breadcrumbs

Clustering

Unsupervised cluster classification

Parameters

Number of cluster

The number of cluster you want to detect and visualize.

Model

Which clustering model type to use:

  • K-Means
    K-means assigns data points to the nearest of centroids, iteratively refining these centroids until convergence.

  • Gaussian Mixture Mode
    GMM represents data using a combination of multiple Gaussian distributions to capture diverse cluster shapes in multidimensional space.

Pretreatments

In top-down applied order ⬇️ , meaning Savitzky-Golay is applied first then Derivative and so on.

  • Savitzky-Golay

  • Derivative

  • SNV (Standard Normal Variate)

  • Logarithm

  • Center

  • UV (Unit Variance) scaling

For more on each type of pretreatment see: Pretreatments