This paper shows that the implicit bias of gradient descent on linearly separable data is exactly characterized by the optimal solution of a dual optimization problem given by a smoothed margin, even for general losses. This is in contrast to prior results, which are often tailored to exponentially-tailed losses. For the exponential loss specifically, with $n$ training examples and $t$ gradient descent steps, our dual analysis further allows us to prove an $O(\ln(n)/\ln(t))$ convergence rate to the $\ell_2$ maximum margin direction, when a constant step size is used. This rate is tight in both $n$ and $t$, which has not been presented by prior work. On the other hand, with a properly chosen but aggressive step size schedule, we prove $O(1/t)$ rates for both $\ell_2$ margin maximization and implicit bias, whereas prior work (including all first-order methods for the general hard-margin linear SVM problem) proved $\widetilde{O}(1/\sqrt{t})$ margin rates, or $O(1/t)$ margin rates to a suboptimal margin, with an implied (slower) bias rate. Our key observations include that gradient descent on the primal variable naturally induces a mirror descent update on the dual variable, and that the dual objective in this setting is smooth enough to give a faster rate.
We study the implicit bias of generic optimization methods, such as mirror descent, natural gradient descent, and steepest descent with respect to different potentials and norms, when optimizing underdetermined linear regression or separable linear classification problems. We explore the question of whether the specific global minimum (among the many possible global minima) reached by an algorithm can be characterized in terms of the potential or norm of the optimization geometry, and independently of hyperparameter choices such as step-size and momentum.
Limitation of Characterizing Implicit Regularization by Data-independent Functions
University of Illinois - Urbana Champaign（伊利诺伊大学厄巴纳-香槟分校）；Shanghai Jiao Tong University（上海交通大学）；Shanghai Artificial Intelligence Laboratory（上海人工智能实验室）；Shanghai Center for Brain Science and Brain-Inspired Technology（上海脑科学与类脑研究中心）
Consider the problem of approximating the optimal policy of a Markov decision process (MDP) by sampling state transitions. In contrast to existing reinforcement learning methods that are based on successive approximations to the nonlinear Bellman equation, we propose a Primal-Dual $π$ Learning method in light of the linear duality between the value and policy. The $π$ learning method is model-free and makes primal-dual updates to the policy and value vectors as new data are revealed. For infinite-horizon undiscounted Markov decision process with finite state space $S$ and finite action space $…
MIT CSAIL（麻省理工学院计算机科学与人工智能实验室）；IBM Research（IBM研究院）；MIT-IBM Watson AI Lab（麻省理工学院-IBM沃森人工智能实验室）
Wasserstein barycenters provide a geometrically meaningful way to aggregate probability distributions, built on the theory of optimal transport. They are difficult to compute in practice, however, leading previous work to restrict their supports to finite sets of points. Leveraging a new dual formulation for the regularized Wasserstein barycenter problem, we introduce a stochastic algorithm that constructs a continuous approximation of the barycenter. We establish strong duality and use the corresponding primal-dual relationship to parametrize the barycenter implicitly using the dual potential…
The Implicit Bias of Steepest Descent with Mini-batch Stochastic Gradient
School of Statistics, Renmin University of China（中国人民大学统计学院）；School of Computing and Data Science, The University of Hong Kong（香港大学计算与数据科学学院）；School of Computing and Data Science & Institute of Data Science, The University of Hong Kong（香港大学计算与数据科学学院及数据科学研究所）
We study non-convex subgradient flows for training two-layer ReLU neural networks from a convex geometry and duality perspective. We characterize the implicit bias of unregularized non-convex gradient flow as convex regularization of an equivalent convex model. We then show that the limit points of non-convex subgradient flows can be identified via primal-dual correspondence in this convex optimization problem. Moreover, we derive a sufficient condition on the dual variables which ensures that the stationary points of the non-convex objective are the KKT points of the convex objective, thus pr…
We propose a new policy gradient method, named homotopic policy mirror descent (HPMD), for solving discounted, infinite horizon MDPs with finite state and action spaces. HPMD performs a mirror descent type policy update with an additional diminishing regularization term, and possesses several computational properties that seem to be new in the literature. We first establish the global linear convergence of HPMD instantiated with Kullback-Leibler divergence, for both the optimality gap, and a weighted distance to the set of optimal policies. Then local superlinear convergence is obtained for bo…
University of Basel（巴塞尔大学）；ELLIS Institute Tübingen（埃利斯研究所蒂宾根分所）；MPI for Intelligent Systems Tübingen（蒂宾根智能系统马克斯·普朗克研究所）；AI Center（人工智能中心）；Yahoo! Research（雅虎研究院）；University of Oslo（奥斯陆大学）
University of Adelaide（阿德莱德大学）；Australian National University（澳大利亚国立大学）
Characterizing the remarkable generalization properties of over-parameterized neural networks remains an open problem. In this paper, we promote a shift of focus towards initialization rather than neural architecture or (stochastic) gradient descent to explain this implicit regularization. Through a Fourier lens, we derive a general result for the spectral bias of neural networks and show that the generalization of neural networks is heavily tied to their initialization. Further, we empirically solidify the developed theoretical insights using practical, deep networks. Finally, we make a case …
Adversarial examples, crafted by adding imperceptible perturbations to natural inputs, can easily fool deep neural networks (DNNs). One of the most successful methods for training adversarially robust DNNs is solving a nonconvex-nonconcave minimax problem with an adversarial training (AT) algorithm. However, among the many AT algorithms, only Dynamic AT (DAT) and You Only Propagate Once (YOPO) guarantee convergence to a stationary point. In this work, we generalize the stochastic primal-dual hybrid gradient algorithm to develop semi-implicit hybrid gradient methods (SI-HGs) for finding station…
This work designs an image restoration deep network relying on unfolded Chambolle-Pock primal-dual iterations. Each layer of our network is built from Chambolle-Pock iterations when specified for minimizing a sum of a $\ell_2$-norm data-term and an analysis sparse prior. The parameters of our network are the step-sizes of the Chambolle-Pock scheme and the linear operator involved in sparsity-based penalization, including implicitly the regularization parameter. A backpropagation procedure is fully described. Preliminary experiments illustrate the good behavior of such a deep primal-dual networ…
