I have a spatial grid with points having multidimensional data. What is the best way to do clustering or segmentation on this? To reiterate, each data point has a spatial coordinate and multi-dimensinal data associated (Multiple features), I want to make use of these features but also taking into account the spatial information.
I have tried building a graph with each spatial point connected to the nearest 5 spatial points with a directed weight as some mathematical metric (Jaccard similarity), then running community detection algorithms as Louvain/Leiden without success.
2.1m questions
2.1m answers
60 comments
57.0k users