Discriminant analysis classifies observations by modeling the distribution of predictors within each class. Linear discriminant analysis assumes class-conditional multivariate normal distributions with class-specific means and a common nonsingular covariance matrix, producing linear decision boundaries. Allowing different nonsingular covariance matrices gives quadratic discriminant analysis.
Discriminant Analysis
See also
Classification, Covariance Matrix, Multivariate AnalysisExplore with Wolfram|Alpha
References
Anderson, T. W. An Introduction to Multivariate Statistical Analysis, 3rd ed. New York: Wiley, 2003.Cite this as:
Weisstein, Eric W. "Discriminant Analysis." From MathWorld--A Wolfram Resource. https://mathworld.wolfram.com/DiscriminantAnalysis.html