Optimization
Regularized Gradient Descent Ascent for Two-Player Zero-Sum Markov Games
Exploiting this property, Daskalakis et al. [2020] builds on the theory of Lin et al. [2020a] and shows
Local Latent Space Bayesian Optimization over Structured Inputs
Here the DAE dramatically simplifies the search space by mapping inputs into a continuous latent space where familiar Bayesian optimization tools can be more readily applied. Despite this simplification, the latent space typically remains high-dimensional.
Supplementary Document On Image Segmentation With Noisy Labels Characterization and Volume Properties of the Optimal Solutions to Accuracy and Dice
Author is also affiliated with RaySearch Laboratories. Here is an incomplete list. Our proof consist of two parts. There are two cases to consider. There are two cases to consider.
A Quantitative Geometric Approach to Neural-Network Smoothness
We implement the algorithms induced from this quantitative geometric approach, which are based on semidefinite programming (SDP).
Adv Attribute Inconspicuous and Transferable Adversarial Attack on Face Recognition Supplementary Material
StyleGAN [1] and the proposed Adv-Attribute attack. During training, the proposed important-aware attribute selection can choose the optimal attribute for the different pairs of target faces and source faces. When attacking the same target face, diverse source faces choose different attributes in each step. Lemma 1. Suppose the overall training loss Do the main claims made in the abstract and introduction accurately reflect the paper's If you ran experiments... (a) Did you include the code, data, and instructions needed to reproduce the main experimental results (either in the supplemental material or as a URL)? [Y es] (b) Did you specify all the training details (e.g., data splits, hyperparameters, how they Did you report error bars (e.g., with respect to the random seed after running experiments multiple times)? Did you include the total amount of compute and the type of resources used (e.g., type Did you include any new assets either in the supplemental material or as a URL? [N/A] Did you discuss whether and how consent was obtained from people whose data you're If you used crowdsourcing or conducted research with human subjects... (a)
Finite-Time Last-Iterate Convergence for Learning in Multi-Player Games
The standard metric to quantify an online learning algorithm's performance is the regret.
main-neurips22
Since weak learners perform only marginally better than random guesses, such subroutines constitute a weaker assumption than the availability of an accurate supervised learning oracle.