the software I am using has libnado in the backend for nn. It uses the createKDTreeLinearHeap. Unfortunatly the search becomes very slow with higher dimensions. I know that libnado is not made for higher dims but is there a way to tune it a little bit? Like another type of tree?
Hey,
the software I am using has libnado in the backend for nn. It uses the createKDTreeLinearHeap. Unfortunatly the search becomes very slow with higher dimensions. I know that libnado is not made for higher dims but is there a way to tune it a little bit? Like another type of tree?