Answers are grounded in this paper and cite their sources.
is said to be consistent, if an element xEK satisfying all m inequalities (i.e. a solution of (1)) does exist. Otherwise (1) is said to be inconsistent. A system (1) is said to be irreducibly inconsistent, if it is inconsistent and if every proper subsystem of (1) is consistent. The convex functions fi, f2, , fm are said to be linearly independent, if no linear combination ZJ-,1 Xif with real coefficients Xi, not all zero, can remain 20 throughout K. In the special case when K =X and when the fi's are linear forms on X, this definition of linear independence agrees with the usual one. Indeed, a linear form can remain >0 throughout the entire vector space X only when it is identically zero. Incidentally, we observe that there exist arbitrarily many linearly independent convex functions even on a one-dimensional convex set. For example, the m convex functions f,(x) =xi-1/(i+1) (1_i_m) on the unit interval 0?x?1 are linearly independent for any natural number m. In fact, if some m real numbers Xi satisfy the reIation ET1 Xi(xi-1/(i+1)) 2O0 for 0?x?<1, then since we have f['[E!1 1Xi(xi-1/(i+1))]dx=0, the polynomial E-1 X,(xi-1/(i+1)) must vanish identically and therefore all X, =0. The purpose of this note is to prove the following results concerning a system (1) of inequalities with convex functions f, defined on K.
No takes yet. Share an insight, caveat, or question.
Fan et al. (1957) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: