Revision as of 16:18, 15 September 2008 edit Trevorgoodchild (talk \| contribs) 294 edits the previous statement was somewhat misleading -- M must be P.D. in order for the objective to be convex in the first place; this is not a special requirement of Lemke's or Dantzig's algorithms ← Previous edit		Revision as of 19:36, 16 September 2008 edit undo Trevorgoodchild (talk \| contribs) 294 edits clarifying previously truncated sentence Next edit →
Line 18: Indeed, these constraints ensure that ''f'' is always non-negative, so that it attains a minimum of 0 at '''x''' if and only if '''x''' solves the linear complementarity problem. ~~Although~~If ~~any~~'''M''' ~~quadratic~~is ~~programming~~[[Positive-definite matrix\|positive definite]], any algorithm for solving (convex) [[Quadratic programming\|QPs]] can of course be used to solve an LCP. However, there also exist more efficient, specialized algorithms, such as [[Lemke's algorithm]] and [[Dantzig's algorithm]]. ==See also==

Linear complementarity problem: Difference between revisions