A Faster Algorithm for Quasi-convex Integer Polynomial Optimization

We present a faster exponential-time algorithm for integer optimization over quasi-convex polynomials. We study the minimization of a quasi-convex polynomial subject to s quasi-convex polynomial constraints and integrality constraints for all variables. The new algorithm is an improvement upon the best known algorithm due to Heinz (Journal of Complexity, 2005). A lower time complexity is reached through applying a stronger ellipsoid rounding method and applying a recent advancement in the shortest vector problem to give a smaller exponential-time complexity of a Lenstra-type algorithm. For the bounded case, our algorithm attains a time-complexity of s (r l M d)^{O(1)} 2^{2n\log_2(n) + O(n)} when M is a bound on the number of monomials in each polynomial and r is the binary encoding length of a bound on the feasible region. In the general case, s l^{O(1)} d^{O(n)} 2^{2n\log_2(n)}. In each we assume d>=2 is a bound on the total degree of the polynomials and l bounds the maximum binary encoding size of the input.

Article

Download

View A Faster Algorithm for Quasi-convex Integer Polynomial Optimization