Gradient Descent Converges to Minimizers

نویسندگان

  • Jason D. Lee
  • Max Simchowitz
  • Michael I. Jordan
  • Benjamin Recht
چکیده

We show that gradient descent converges to a local minimizer, almost surely with random initialization. This is proved by applying the Stable Manifold Theorem from dynamical systems theory.

برای دانلود رایگان متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Gradient Descent Only Converges to Minimizers

We show that gradient descent converges to a local minimizer, almost surely with random initialization. This is proved by applying the Stable Manifold Theorem from dynamical systems theory.

متن کامل

Gradient Descent Only Converges to Minimizers: Non-Isolated Critical Points and Invariant Regions

We prove that the set of initial conditions so that gradient descent converges to strict saddle points has (Lebesgue) measure zero, even for non-isolated critical points, answering an open question in [1].

متن کامل

Stabilizing Adversarial Nets with Prediction Methods

Adversarial neural networks solve many important problems in data science, but are notoriously difficult to train. These difficulties come from the fact that optimal weights for adversarial nets correspond to saddle points, and not minimizers, of the loss function. The alternating stochastic gradient methods typically used for such problems do not reliably converge to saddle points, and when co...

متن کامل

Stabilizing Adversarial Nets with Prediction Methods

Adversarial neural networks solve many important problems in data science, but are notoriously difficult to train. These difficulties come from the fact that optimal weights for adversarial nets correspond to saddle points, and not minimizers, of the loss function. The alternating stochastic gradient methods typically used for such problems do not reliably converge to saddle points, and when co...

متن کامل

Stabilizing Adversarial Nets With Prediction Methods

Adversarial neural networks solve many important problems in data science, but are notoriously difficult to train. These difficulties come from the fact that optimal weights for adversarial nets correspond to saddle points, and not minimizers, of the loss function. The alternating stochastic gradient methods typically used for such problems do not reliably converge to saddle points, and when co...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

عنوان ژورنال:
  • CoRR

دوره abs/1602.04915  شماره 

صفحات  -

تاریخ انتشار 2016