Dual form reduction in the atomic optimization method (Q2399760)

In this work, the author studies one dimensional non-convex optimization problems with polynomial objective function and polynomial matrix inequality constraints for which they construct linear relaxations. The dual problems to the linear relaxations are then investigated, namely certain transformation is provided which shows that such problems are compatible with the atomic optimization method. Some reduction of the number of atoms is also given. Numerous comments allow for better understanding of the author's approach which is supplied with worked examples.