I am PhD Candidate in the Department of Management Science and Engineering at Stanford in the Operations Research group. My advisor is Yinyu Ye. Currently, I focus on developing efficient algorithms for finding local optimum to continuous nonconvex optimization problems. Roughly speaking, my research is split into constrained optimization where I work on interior point methods and unconstrained optimization where I focus on first order methods. However, my broad areas of interests include market design, integer programming, machine scheduling, and machine learning. Feel free to contact me with any questions. My email is ohinder at stanford dot edu. |

Grouped by topic and author lists are alphabetical.

This work studies interior point methods for finding KKT points of the problem subject to . We assume and are twice differentiable functions. We develop theory and implementations. The work consists of the following papers.

A one-phase interior point method for nonconvex optimization.

Oliver Hinder, Yinyu Ye.

Submitted to mathematical programming.

Code. Slides.

On the behavior of Lagrange multipliers in convex and non-convex infeasible interior point methods.

Gabriel Haeser, Oliver Hinder, Yinyu Ye.

Submitted to mathematical programming.

This work studies the worst-case runtime of first-order methods for finding points with under the assumptions that the function derivatives are not changing too quickly (Lipschitz). The work consists of the following papers.

Accelerated Methods for Non-Convex Optimization.

Yair Carmon, John Duchi, Oliver Hinder and Aaron Sidford.

To appear in SIAM journal on optimization.

‘Convex Until Proven Guilty’: Dimension-Free Acceleration of Gradient Descent on Non-Convex Functions.

Yair Carmon, John Duchi, Oliver Hinder and Aaron Sidford.

ICML 2017.

Slides. Video of talk at ICML.

Lower Bounds for Finding Stationary Points I.

Yair Carmon, John Duchi, Oliver Hinder and Aaron Sidford.

Submitted to mathematical programming.

Lower Bounds for Finding Stationary Points II: First-Order Methods.

Yair Carmon, John Duchi, Oliver Hinder and Aaron Sidford.

Submitted to mathematical programming.

For a brief overview of the lower bounds see our NIPS 2017 workshop paper.

A novel integer programming formulation for scheduling with family setup times on a single machine to minimize maximum lateness.

Oliver Hinder, Andrew Mason.

European Journal of Operational Research, 2017.