方法对比

	典型相关分析	偏最小二乘回归 (PLS)
领域	统计学	机器学习
方法族	Latent structure	Machine learning
起源年份	1936	1975
提出者	Harold Hotelling	Herman Wold; popularized by Svante Wold in chemometrics
类型	Multivariate linear dimension reduction and association	Supervised latent-variable regression
开创性文献	Hotelling, H. (1936). Relations between two sets of variates. Biometrika, 28(3–4), 321–377.	Wold, S., Sjöström, M., & Eriksson, L. (2001). PLS-regression: a basic tool of chemometrics. Chemometrics and Intelligent Laboratory Systems, 58(2), 109–130.
别名	CCA, canonical variate analysis, canonical analysis, multiple canonical correlation	PLS regression, projection to latent structures, PLSR, kısmi en küçük kareler
相关	4	3
摘要	Canonical Correlation Analysis (CCA) is a multivariate statistical method that identifies pairs of linear combinations — one from each of two variable sets — such that the correlation between each pair is maximised. Introduced by Harold Hotelling in his landmark 1936 Biometrika paper, CCA provides the most general linear framework for studying the association between two multivariate batteries of measurements, and many classical procedures (multiple regression, MANOVA, discriminant analysis) are special cases of it.	Partial least squares regression predicts a response from many, often highly collinear predictors by projecting them onto a small set of latent components — but, unlike principal components regression, it chooses those components to maximize their covariance with the response, not just the variance of the predictors. This supervised dimension reduction makes PLS a workhorse in chemometrics, spectroscopy, and other wide-data settings where predictors vastly outnumber observations.