A Classical Math Problem Gets Pulled Into the Modern World | Quanta Magazine
A century ago, the great mathematician David Hilbert posed a probing question in pure mathematics. A recent advance in optimization theory is bringing Hilbert’s
Hasnain says:
Interesting read that covers some amount of math and modern robotics/ simulations.
Wish they went into a bit more technical detail but I guess I can read the paper for that.
“But semidefinite programming has a big limitation: It is slow on large problems and can’t handle many of the most complicated polynomials researchers really care about. Semidefinite programming can be used to find a sum of squares decomposition for polynomials with a handful to about a dozen variables raised to powers no higher than about 6. The polynomials that characterize complex engineering problems — like how to ensure a humanoid robot stays on its feet — can involve 50 or more variables. A semidefinite program could chew on that kind of polynomial until the end of time and still not return a sum of squares answer.”
Posted on 2019-11-10T21:10:18+0000