Revision as of 18:52, 23 June 2015 edit Loraof (talk \| contribs) Extended confirmed users 22,850 edits →Nonparametric multiplicative regression{{anchor\|Multiplicative}}: ce ← Previous edit		Revision as of 20:41, 13 July 2015 edit undo Qwertyus (talk \| contribs) Extended confirmed users 31,640 edits →Penalized least squares: removing this section: not specific to non-parametric regression, and the relation is not explicated at all Next edit →
Line 7: [[Smoothing splines]] have an interpretation as the posterior mode of a Gaussian process regression. ~~==Penalized least squares==~~ ~~===Quadratic penalties===~~ ~~Here the penalty term is proportional to the [[RKHS]] norm of the vector of regression coefficients.~~ ~~===Lasso===~~ ~~{{Main\|Lasso (statistics)}}~~ ~~The lasso is a form of penalized least squares, where the penalty term is proportional to the L1 norm of the vector of regression coefficients.~~ ==Kernel regression==

Nonparametric regression: Difference between revisions