Research map / Optimization / Optimizers
Statistical Learning and Inverse Problems: A Stochastic Gradient Approach
Yuri Fonseca, Yuri Saporito
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 20 spline degrees of freedom instead of 10 (ablation) | ML-SGD with spline base learner · Bitcoin darknet classification | avg accuracy: 0.73 -> 0.79 | Improved |
| use 20 spline knots instead of 10 (ablation) | FPLR · Bitcoin darknet classification | avg accuracy: 0.74 -> 0.80 | Improved |
| replace spline base learner with regression tree (depth 20) (ablation) | ML-SGD · Bitcoin darknet classification | avg accuracy: 0.79 -> 0.78 | No clear effect |
| plain SGD-SIP vs ML-SGD with spline k=20 (ablation) | SGD-SIP · Bitcoin darknet classification | avg accuracy: 0.79 -> 0.80 | No clear effect |
| SGD-SIP vs tailored FPLR (k = 20) | SGD-SIP · Bitcoin darknet classification | avg accuracy: 0.80 -> 0.80 | No clear effect |
| smoothing-spline base learners in place of raw stochastic gradients | ML-SGD · Synthetic FLR and deconvolution | MSE / smoothness | Mixed |
| compare SGD-SIP/ML-SGD with Landweber iteration | SGD-SIP · Synthetic FLR (sine) | MSE | No clear effect |
| compare ML-SGD with tailored PFLR on smooth sine target | ML-SGD · Synthetic FLR (sine) | MSE | Hurt |
| use a non-smooth step target function | PFLR advantage over SGD methods · Synthetic FLR (step) | MSE | Hurt |
| use all samples at every iteration instead of one | SGD-SIP / ML-SGD · Synthetic FLR (step) | MSE | Improved |
Closest papers by meaning
- Stochastic Approximate Gradient Descent via the Langevin Algorithm (AAAI 2020 · similarity 0.85)
- Painless Stochastic Gradient: Interpolation, Line-Search, and Convergence Rates (NeurIPS 2019 · similarity 0.84)
- Sampling from Gaussian Process Posteriors using Stochastic Gradient Descent (NeurIPS 2023 · similarity 0.84)
- Online Stochastic Gradient Descent with Arbitrary Initialization Solves Non-smooth, Non-convex Phase Retrieval (JMLR 2023 · similarity 0.84)
- A Guide Through the Zoo of Biased SGD (NeurIPS 2023 · similarity 0.84)
- Demystifying SGD with Doubly Stochastic Gradients (ICML 2024 · similarity 0.84)
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.
