I am trying to cluster by dataset with mixed features using k-means. As a distance metric, I am using Gower's Dissimilarity. I want to ask 2 things:
-Is k-means an appropriate algorithm that can accept Gower's matrix results as an input? or how can I use the output of Gower' matrix as an input of another clustering algorithm?
-Can I apply PCA or Kernel PCA on the results of Gower's Discance to decrease dimensionality so that I can visualize the results?