University of Castilla-La Mancha AI researcher bridging Bayesian networks, genetic algorithms, and federated learning
4
papers here
4
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
28
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—
No output score is available yet. This axis would measure publication volume and trajectory; the snapshot holds raw signals — 23 works in the last three years and yearly counts from 1993 to 2026 — but no computed dimension score.
Impact—
No impact score is available yet. This axis would capture citation-based influence; raw aggregates exist (h-index 28, i10-index 75) but no normalized impact dimension has been computed, so nothing is scored or estimated here.
Focus0/100
The focus axis scores 0, the lowest computed value. The topic distribution explains it: 89 works sit in Bayesian modeling and causal inference, but sizable clusters in data mining algorithms (53), data mining applications (42), metaheuristic optimization (35) and ML classification (27) spread the record across four field labels weighted 0.25 each — breadth, not specialization.
Collaboration39/100
The collaboration axis scores 39, comparatively low. The recent window records only 6 distinct coauthors, and all four cataloged papers place him in a middle author position with no corresponding role. Recurring partners — Pablo Torrijos, José M. Puerta, Juan C. Alfaro, Fabio Stella and Juan A. Aledo — suggest a small, stable, long-running circle rather than a broad shifting network.
Activity93/100
The activity axis scores 93, among the profile's strongest signals. Yearly work counts run continuously from 1993 (2 works) through 2026 (6 works), peaking at 15 in both 2011 and 2016, and the last three years contribute 23 works, with 8 in 2023, 10 in 2024, 7 in 2025 and 6 in 2026 — sustained, near-uninterrupted productivity across more than three decades.
Reproducibility89/100
The reproducibility axis scores 89, computed as reproMean across the 4 Evolve-reviewed papers. Three carry a high reproducibility-readiness rating and one medium; the treewidth-limited fusion paper explicitly released code. Reviews credit artifact availability while flagging self-baseline comparisons and missing significance tests — a strong open-science signal with acknowledged methodological gaps.
Research summary
José A. Gámez is a long-standing artificial intelligence researcher at the University of Castilla-La Mancha whose OpenAlex record spans 1993 to 2026. His work centers on probabilistic graphical models — Bayesian networks and their classifier variants — combined with evolutionary computation; the platform's topic model attributes 89 works to Bayesian modeling and causal inference, with further clusters in data mining, metaheuristic optimization, and machine learning classification. The snapshot reports an h-index of 28, an i10-index of 75, and 23 works over the last three years, scoring activity at 93 and reproducibility at 89 (reproMean over 4 reviewed papers), while output and impact dimensions remain uncomputed. The four Evolve-reviewed 2026 arXiv papers form a coherent late-career program: FedAnDE, the first federated framework for discriminative Averaged n-Dependence Estimators; two companion papers on treewidth-limited structural fusion of Bayesian networks using genetic algorithms, one with released code; and FedRCD, a federated LiNGAM-style causal discovery method whose review honestly notes that cumulant-based ranking tracks variance rather than causal asymmetry. Collaboration scores a modest 39, consistent with 6 distinct recent coauthors and middle author positions on all cataloged papers, anchored by long-term colleagues such as José M. Puerta and Pablo Torrijos. Overall: a broad, persistent graphical-models researcher now folding federated learning and causal discovery into a three-decade Bayesian agenda.
Collaboration pattern
The collaboration profile is compact: only 6 distinct coauthors in the recent window and a collaboration score of 39. All four cataloged papers list him in a middle position with no corresponding-author role, pointing to team publications. Recurring partners are Pablo Torrijos, José M. Puerta, Juan C. Alfaro, Fabio Stella and Juan A. Aledo — a blend of long-standing Castilla-La Mancha colleagues and external collaborators such as Stella.
Research keywords
Highlights
FedAnDE: first federated framework for discriminative AnDE classifiers
Treewidth-limited Bayesian network fusion with genetic algorithms
Pre-fusion edge pruning makes consensus Bayesian network fusion tractable
FedRCD: federated causal discovery — and an honest look at cumulant ranking
Timeline
- 1993The bibliometric record opens with 2 works in 1993, the earliest year in the counts-by-year snapshot, marking the start of a record that continues unbroken to 2026.
- 2011Yearly output reaches 15 works, the highest count in the 1993–2026 series — a level matched only in 2016.
- 2016A second peak of 15 works; the surrounding high points (13 works in 2006, 2007 and 2010) confirm a sustained mid-career stretch.
- 2026Four Evolve-reviewed arXiv papers appear: federated AnDE classifiers, two papers on treewidth-limited Bayesian network fusion with genetic algorithms, and federated LiNGAM-style causal discovery.
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 are drawn from an OpenAlex-derived snapshot; output and impact dimension scores are not computed, so no citation-normalized assessment is offered.
- All four cataloged papers are 2026 arXiv preprints; the reproducibility evidence (reproMean 89 over 4 papers) covers only this small sample.
- The profile is name-resolved from public bibliographic data (OpenAlex identity); the snapshot's country code (IT) conflicts with the listed Spanish institution, so affiliation metadata should be treated with caution.
- Topic and field distributions come from automatic classification and may not match the researcher's self-described areas.