Research map / Optimization / Nonconvex and smooth optimization
Three Operator Splitting with Subgradients, Stochastic Gradients, and Adaptive Learning Rates
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use l1 loss instead of l2 loss for matrix completion with salt and pepper noise (ablation) | ADAPTOS · image inpainting and denoising | PSNR: 21.15 dB -> 26.21 dB | Helped |
| l2 loss instead of l1 loss (ablation) | ADAPTOS · image inpainting and denoising | convergence rate | Mixed |
| use prox-operators of regularizers (ADAPTOS) instead of subgradient methods | group-sparse regularized neural network · MNIST | sparsity: 93% weights zero, 68% neurons inactive | Helped |
| ADAPTOS versus SGD/Adam on neural network | ADAPTOS · MNIST fully connected network | train and test accuracy | Helped |
| apply ADAPTOS to more complex models such as ResNet | ADAPTOS · ResNet preliminary experiments | performance | Hurt |
| ADAPTOS versus TOS-LS and PDHG-LS on smooth convex problems | ADAPTOS · overlapping group lasso, TV deblurring, sparse low-rank recovery | convergence speed | Mixed |
Closest papers by meaning
- Adaptive Three Operator Splitting (ICML 2018 · similarity 0.91)
- Stochastic Three-Composite Convex Minimization (NeurIPS 2016 · similarity 0.90)
- Three Operator Splitting with a Nonconvex Loss Function (ICML 2021 · similarity 0.89)
- Universality of AdaGrad Stepsizes for Stochastic Optimization: Inexact Oracle, Acceleration and Variance Reduction (NeurIPS 2024 · similarity 0.88)
- On the Almost Sure Convergence of the Stochastic Three Points Algorithm (ICLR 2025 · similarity 0.87)
- Universal Gradient Methods for Stochastic Convex Optimization (ICML 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.