{"id":"arxiv/2611.23414","url":"https://arxiv.org/abs/2611.23414","tags":["cs.LG","math.FA","math.NA"],"score":0,"title":"Incremental Roseland: Batch-wise Landmark Diffusion via Kernelized Sequential Karhunen--Loève Update","domain":"cs.LG","source":"arxiv","authors":["Jeanie Schreiber"],"reviews":0,"subject":"cs.LG","summary":"Landmark diffusion (Roseland) computes a diffusion-map-style spectral embedding from the singular value decomposition of a degree-normalized data-to-landmark kernel matrix, at a cost far below that of full diffusion maps. It assumes, however, that all data are available at once. We introduce incremental Roseland (iRoseland), which casts the landmark affinity as a row-growing matrix whose SVD can be updated with a kernelized Sequential Karhunen--Loève (SKL) scheme. Degre correction is handled by a first order expansion about a reference mean kernel row and so that the landmark-side eigenbasis…","category":"research","comments":0,"disputed":false,"division":"cs","mentions":0,"discussed":0,"published":"2026-11-17","reactions":{}}