Graph signal processing researcher: stability-aware graph neural networks and generative modeling of graph signals
2
papers here
2
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
31
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—
Output score not computed: no verified total publication count is available in this snapshot. This axis would measure cumulative works and their distribution across the career, complementing the h-index (31) and i10-index (58) already recorded.
Impact—
Impact score not computed: citation totals and per-paper citation distributions are unavailable in the current data. This axis would quantify citation-based influence, including field-normalized citation rates and highly cited works, beyond the h-index (31) and i10-index (58).
Focus0/100
The platform scores focus at 0, tracking the raw field distribution: output splits evenly (0.5/0.5) between 'signal processing / graph machine learning' and 'graph signal processing (generative machine learning)'. Read against topic counts, the portfolio is nonetheless concentrated — advanced graph neural networks (87 works) and sparse and compressive sensing (70) dominate, followed by complex network analysis (33), graph theory and algorithms (32), and statistical methods and inference (29).
Collaboration36/100
Collaboration scores 36. The snapshot lists 5 distinct recent coauthors against 49 works over the last three years — a compact core rather than a large rotating network. Recurring partners include Saurabh Sihag, Andrea Cavallo, Elvin Isufi, Alejandro Ribeiro, and Martin Schmidt, and both reviewed 2026 papers are multi-author, with the researcher in last and middle author positions.
Activity100/100
Activity scores 100, the maximum. The yearly counts show 49 works over 2024–2026 (15, 20, 14), the most productive span in a record reaching back to 2007, when output was a single work. It exceeds earlier peaks of 13 in 2016 and 14 in both 2018 and 2022, indicating accelerating rather than tapering output.
Reproducibility55/100
Reproducibility scores 55, computed from 2 reviewed papers, each rated medium readiness. The 2026 generative-filter study ships code with sound theory but small-scale validation; the covariance/PCA tutorial makes sound claims with illustrative but thin experiments. With only two papers assessed, the score is indicative and should firm up as the catalog grows.
Research summary
Gonzalo Mateos (Universidad Rey Juan Carlos, Spain) is a graph signal processing researcher whose OpenAlex-resolved record spans two decades, from a single work in 2007 to an annual peak of 20 in 2025. The profile carries an h-index of 31 and an i10-index of 58, with 49 works over the last three years (2024–2026: 15, 20, 14) — the most productive span on record and the basis for a maximum activity score. Thematic mass sits in advanced graph neural networks (87 works) and sparse and compressive sensing (70), with smaller clusters in complex network analysis (33), graph theory and algorithms (32), and statistical methods and inference (29); the platform splits the field evenly (0.5/0.5) between signal processing/graph machine learning and graph signal processing/generative machine learning. Two 2026 arXiv papers are reviewed in the Evolve catalog: 'Stable Filters for Generative Modeling of Graph Signals,' which derives Wasserstein stability bounds for graph-aware generative dynamics plus an LP-based stable-filter design with code (last author), and 'Learning with Covariance Matrices,' a tutorial extending PCA to covariance neural networks with stability and multiscale transferability theory (middle author). Both are rated medium reproducibility readiness, averaging 55 across two papers. Collaboration scores 36, with 5 distinct recent coauthors and recurring partners including Saurabh Sihag, Andrea Cavallo, Elvin Isufi, Alejandro Ribeiro, and Martin Schmidt.
Collaboration pattern
The profile lists 5 distinct recent coauthors against 49 works over the last three years — a compact core rather than a large rotating network, consistent with the moderate collaboration score (36). Recurring names include Saurabh Sihag, Andrea Cavallo, Elvin Isufi, Alejandro Ribeiro, and Martin Schmidt. In the two reviewed 2026 papers the researcher appears as last author on the generative-filter study and middle author on the PCA–covariance tutorial, neither as corresponding author.
Research keywords
Highlights
Wasserstein-stable filters for graph-signal generative models
Tutorial uniting covariance matrices, PCA, and graph learning
Most productive stretch on record
Yearly counts show 49 works across 2024–2026 (15, 20, 14), with 2025's 20 works the highest annual total in a record reaching back to 2007 — the basis for the maximum activity score.
Timeline
- 2007Earliest year in the bibliometric record, with a single work — the starting point of a two-decade publishing trace.
- 2016Output steps up to 13 works after 4 in 2015, opening a growth phase that also reaches 14 works in 2018 and 2022.
- 2024The most productive three-year span in the record begins, totaling 49 works across 2024–2026 (15, 20, 14).
- 2025Annual peak: 20 works, the highest yearly count in the record.
- 2026Two reviewed arXiv papers appear: 'Stable Filters for Generative Modeling of Graph Signals' (last author) and 'Learning with Covariance Matrices' (middle author).
Papers in the Evolve catalog
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.
- Bibliometrics come from an OpenAlex snapshot; 'Gonzalo Mateos' is a common Spanish name, so verify identity against ORCID 0000-0002-9847-6298 before citing.
- Output and impact dimension scores are missing because citation totals are unavailable; only the h-index (31) and i10-index (58) can be quoted.
- The reproducibility score (55) rests on only 2 reviewed papers and may shift as the Evolve catalog grows.
- Field labels are platform-assigned and split evenly (0.5/0.5) between two overlapping graph-related categories.