KnowraPrincipal component analysisLinked fromLinked fromThe 28 pages that link to Principal component analysis, each with the reason it gives.All 28Broader topic 2Related 23Compared with 3Orthonormal basisCompared with: Its finite orthogonal coordinates resemble basis expansions but are chosen from data variance.Factor analysisCompared with: Unlike factor analysis, it represents total observed variance rather than modeling shared variance and errors.Johnson–Lindenstrauss lemmaCompared with: Unlike variance-maximizing PCA, random projections need not inspect the data to preserve pairwise distances.