KnowraLeast squaresLinked fromLinked fromThe 68 pages that link to Least squares, each with the reason it gives.All 68Related 52Narrower topic 4Compared with 12Maximum likelihood estimationCompared with: Under Gaussian errors with constant variance, least squares and maximum likelihood give the same coefficients.Polynomial interpolationCompared with: Unlike interpolation, it can tolerate inconsistent data without passing through every point.Principle of least actionCompared with: Unlike stationary action, it is an inference method rather than a law of motion.System of linear equationsCompared with: It provides a best-fit alternative when an overdetermined system has no exact solution.Robust statisticsCompared with: Squaring residuals gives extreme observations disproportionate influence compared with robust alternatives.Rational approximationCompared with: Least-squares fitting minimizes aggregate error rather than rational approximation’s possible worst-case error.Method of momentsCompared with: Regression parameters can be fitted by minimizing residuals instead of matching distribution moments.Matrix inverseCompared with: Normal equations introduce an inverse, though stable algorithms often avoid computing it directly.Cramer's ruleCompared with: It addresses systems that are not square, outside Cramer's rule's setting.Lagrange interpolationCompared with: Unlike exact interpolation, least squares generally permits residuals at the supplied data points.Infimum and supremumCompared with: Optimization often seeks an attained minimum, unlike an infimum that may not be attained.Rouché–Capelli theoremCompared with: When ranks differ, least squares finds an approximation rather than an exact solution.