Sparse solutions to random standard quadratic optimization problems (Q378104)

The authors refer to a standard quadratic optimization problem of minimizing a quadratic form over the standard simplex. The work is an initial attempt to establish the existence of sparse optimal globally solutions for a class of random matrices. Specially, when the upper triangular elements of the matrix, involved in the quadratic form, are identically independently distributed, with a uniform or exponential distribution. It has been proven that the probability that there exists a globally optimal solution with at least k nonzero elements decays exponentially (in \(k\)).