Jiaqi Leng

Abstract We initiate the study of quantum algorithms for escaping from saddle points with provable guarantee. Given a function $f:\R^{n}\to\R$, our quantum algorithm outputs an $\epsilon$-approximate local minimum using $\tilde{O}(\log^{2} n/\epsilon^{1.75})$ queries to the quantum evaluation oracle (i.e., the zeroth-order oracle). Compared to the classical state-of-the-art algorithm by Jin et al.~with $\tilde{O}(\log^{6} n/\epsilon^{1.75})$ queries to the gradient oracle (i.e., the first-order oracle), our quantum algorithm is polynomially better in terms of $n$ and matches its complexity in terms of $1/\epsilon$. Our quantum algorithm is built upon two techniques: First, we replace the classical perturbations in gradient descent methods by simulating quantum wave equations, which constitutes the polynomial speedup in $n$ for escaping from saddle points. Second, we show how to use a quantum gradient computation algorithm due to Jordan to replace the classical gradient queries in nonconvex optimization by quantum evaluation queries with the same complexity, extending the same result from convex optimization due to van Apeldoorn et al. and Chakrabarti et al. Finally, we also perform numerical experiments that support our quantum speedup.