So sánh phương pháp

	Phân tích hỗn hợp siêu phổ	Phân tích ma trận không âm (NMF)
Lĩnh vực	Viễn thám	Học máy
Họ	Machine learning	Latent structure
Năm ra đời	2002	1999
Người khởi xướng	Nirmal Keshava & John Mustard	Lee, D. D. & Seung, H. S.
Loại	Sub-pixel spectral decomposition algorithm	Matrix decomposition with non-negativity constraints
Công trình gốc	Keshava, N., & Mustard, J. F. (2002). Spectral unmixing. IEEE Signal Processing Magazine, 19(1), 44–57.	Lee, D. D., & Seung, H. S. (1999). Learning the parts of objects by non-negative matrix factorization. Nature, 401(6755), 788–791.
Tên gọi khác	Spectral Mixture Analysis, Linear Spectral Unmixing, Blind Source Separation (Hyperspectral), Hiperspektral Ayrıştırma	NMF, NNMF, nonnegative matrix factorization, non-negative matrix approximation
Liên quan	2	4
Tóm tắt	Hyperspectral unmixing is a signal processing technique that decomposes each pixel of a hyperspectral image into a collection of pure material spectra (endmembers) and their corresponding fractional abundances. Because sensor resolution often causes multiple land-cover types to co-occupy a single pixel, unmixing recovers sub-pixel compositional information that conventional classification cannot. Keshava and Mustard (2002) provided the foundational signal-processing framework that unified prior geological and remote-sensing work under a rigorous linear mixture model.	Non-negative Matrix Factorization (NMF) is a family of algorithms, introduced by Lee and Seung in their landmark 1999 Nature paper, that decomposes a non-negative data matrix V into the product of two lower-rank non-negative matrices W (basis components) and H (encoding coefficients). Unlike PCA or SVD, the non-negativity constraint forces the algorithm to learn strictly additive, parts-based representations, making the factors directly interpretable as building blocks of the original data.