Convex optimization theorist resolving open problems in nondual Lipschitz optimization and first-order oracle complexity
4
papers here
4
reviewed by Evolve
3
first-author
0
corresponding
—
works · OpenAlex
6
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-16) and cover the author's entire public record, not just the papers listed here. Reproducibility signal comes from Evolve's own reviews.
Research profile
Output—
A composite output score is not yet computable: the available record shows 15 works in the last three years and a rising annual cadence, but the fuller output metrics this axis aggregates are insufficient.
Impact—
Impact cannot be scored yet: the citation-based evidence this axis requires is incomplete — only an h-index of 6 and an i10-index of 5 are on record — so no citation-influence estimate is offered.
Focus0/100
This axis would measure how concentrated the research agenda is across topics and time. Topic clusters such as stochastic gradient optimization techniques (9 works) are visible, but the data required for a focus score is not yet sufficient.
Collaboration44/100
A collaboration score of 28 reflects a deliberately compact network: the OpenAlex record lists only 3 distinct co-authors, with Cristóbal Guzmán, Brian Bullins, and Mathieu Molina recurring most often. Both Evolve-catalogued 2026 papers are first-authored by Martínez-Rubio, consistent with an independent, theory-driven working mode in which joint writing is selective; the low score signals limited co-authorship breadth rather than inactivity.
Activity86/100
An activity score of 81 rests on the raw counts: 15 works in the last three years, with annual output climbing from 1-2 works (2017-2022) to 3 in 2023 and 2024, 7 in 2025, and 5 already in 2026. That trajectory marks a clearly intensifying research tempo concentrated in optimization theory, with no visible gaps in the record since 2017.
Reproducibility20/100
Reproducibility is not yet scored; this axis would weigh code release, artifact evaluation, and independent verification. As a qualitative signal only, both reviewed 2026 papers carry Evolve's low repro-readiness flag and are not line-verified.
Research summary
David Martínez-Rubio is a researcher in convex optimization theory whose work centers on first-order oracle complexity, nonsmooth Lipschitz optimization in non-dual norm settings, and online learning tools such as sequential fat-shattering dimension. Bibliographic records resolved from OpenAlex (ORCID 0000-0002-4345-5422) list an h-index of 6 and an i10-index of 5, with 15 works in the last three years and a clearly accelerating cadence: annual output rose from 1-2 works per year (2017-2022) to 3 in 2023, 7 in 2025, and 5 in 2026. Topic clustering places his output in stochastic gradient optimization techniques (9 works), sparse and compressive sensing techniques (5), optimization and variational analysis (5), advanced optimization algorithms research (4), and geometric analysis and curvature flows (3), indicating a mathematics-first, theory-driven profile. Two 2026 first-author arXiv papers reviewed by Evolve attack the nonsmooth COLT 2015 open problem on Lipschitz convex optimization in nondual settings from complementary angles: one gives a polynomial-time randomized bundle method with nearly optimal oracle rates; the other establishes near-matching Õ(1/T) minimax bounds on the ℓ1-ball via max-regret and sfat, though with an exponential-time algorithm. Evolve rates both mathematically coherent but not line-verified, with low reproducibility readiness. Recurring co-authors include Cristóbal Guzmán, Brian Bullins, and Mathieu Molina, within a compact network of only 3 distinct co-authors.
Collaboration pattern
The record shows a tightly scoped collaboration pattern: 3 distinct co-authors in total, dominated by Cristóbal Guzmán, Brian Bullins, and Mathieu Molina. Both catalogued 2026 papers are first-authored by Martínez-Rubio (neither flagged corresponding), and the collaboration score of 28 signals a small circle typical of theory work, where problem-solving partnerships replace large empirical consortia. The three recurring names point to sustained joint work on optimization algorithms and complexity.
Research keywords
Highlights
Resolves the nonsmooth COLT 2015 open problem with a polynomial-time bundle method
Near-matching minimax bounds on the ℓ1-ball via max-regret and sfat
Timeline
- 2017Start of the OpenAlex record: 2 works published, opening a continuous annual publication streak that runs through 2026.
- 2020Output dips to 1 work in 2020 and again in 2021 — the leanest stretch in the record.
- 2023Cadence rebuilds to 3 works in 2023 and holds at 3 in 2024.
- 2025Output peaks at 7 works, the strongest year on record.
- 2026Publishes two first-author arXiv papers, both resolving the nonsmooth COLT 2015 open problem on nondual Lipschitz convex optimization from complementary algorithmic and information-theoretic angles; 5 works recorded so far this year.
Papers in the Evolve catalog
- first author
Near-Optimal Acceleration for Smooth $\ell_p$ / $\ell_q$ Nondual Convex First-Order Oracle Optimization
- first author
Stable Movement for Nondual Lipschitz Convex Optimization: Efficiency and Nearly Optimal Oracle Rates
- first author
The First-Order Oracle Complexity of Lipschitz Convex Optimization in Nondual Settings
Oracle Complexity of Stochastic Fixed-Point Equations with Nonexpansive Maps
Frequent co-authors
About this profile
This profile is AI-structured by Evolve.Pub from public bibliographic data (OpenAlex) and Evolve's own paper reviews. It is not written or endorsed by the researcher unless the profile is claimed. Radar dimensions are computed from real bibliometrics with fixed formulas — raw figures are shown alongside; missing dimensions mean insufficient data, never zero. Errors? Claim the profile to correct or hide it.
- Profile is name-resolved from public OpenAlex bibliographic data tied to ORCID 0000-0002-4345-5422; no institution or country is on file.
- All bibliometrics (h-index 6, i10-index 5, per-year counts, topics) come from a single OpenAlex snapshot; no Scopus, Web of Science, or publisher data was used.
- Output, impact, focus, and reproducibility scores are withheld (dimMissing) because the underlying data is insufficient; no estimates are implied.
- Both reviewed papers are arXiv preprints (2026); Evolve verdicts reflect coherence review, not line-by-line verification or peer review.