Title:  Shrink-wrapping Trajectories for Linear Programming

Hyperbolic Programming (HP) -minimizing linear objective over an affine subspace intersected with a hyperbolicity cone- is a class of convex optimization problems containing Linear Programming (LP) and its natural relaxations. Based on these relaxations a new Shrink- Wrapping approach for LP has been proposed by Renegar. We analyze Shrink-Wrapping trajectories near a certain invariant set containing LP optimum and contrast the trajectories with the central path on some pathological LP instances.