Dimitris K. Iakovidis
—
Explainable neuro-fuzzy machine learning for predictive maintenance in naval and maritime systems
2
papers here
2
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
—
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-12) 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—
Not yet scored: Evolve has no verified publication-count, venue, or total-output record for this name beyond the two arXiv preprints reviewed. This axis would measure cumulative paper volume, output consistency, and venue spread over time.
Impact—
Not yet scored: no citation counts, h-index, or altmetrics are available in this snapshot, and both reviewed items are recent 2026 preprints. This axis would measure citations, downstream uptake, and influence on the explainable-maintenance literature.
Focus0/100
The computed focus score is 0, but the raw field distribution is nearly unified: 0.5 in predictive maintenance / applied machine learning and 0.5 in machine learning (explainable AI / predictive maintenance) — two labels describing essentially one theme. With only 2 papers in the catalog, the score is unstable; the reviewed work itself shows a tightly concentrated explainable-AI-for-maintenance agenda across maritime and naval domains.
Collaboration28/100
The collaboration score is 28, derived from 3 distinct co-authors — Dionisis Kalogeropoulos, Georgia Sovatzidi, and Panagiotis G. Kalozoumis — across the two reviewed papers. Both are preprints, and both list him in the last author position without corresponding-author status, suggesting a compact, possibly supervisory-team pattern rather than a broad external network.
Activity—
Not yet scored: the snapshot contains only two 2026 arXiv preprints and no multi-year publication series, so publication cadence cannot be computed. This axis would measure output tempo — papers per year, gaps, and sustained research activity over time.
Reproducibility20/100
The reproducibility score is 20, a mean taken over 2 reviewed papers, both rated low in reproducibility readiness by Evolve. The naval-propulsion paper uses a public CODLAG dataset but faces baseline-comparison and possible-leakage concerns; the maritime paper lacks variance statistics. Neither preprint currently signals full code or data release.
Research summary
Dimitris K. Iakovidis is an early-stage applied machine-learning researcher whose reviewed work centers on explainable artificial intelligence for predictive and condition-based maintenance in naval and maritime settings. Two 2026 arXiv preprints in the Evolve catalog — both listing him as last author — share one methodological signature: a neural or neuro-fuzzy predictor distilled into interpretable fuzzy decision trees, supplemented by saliency-based feature attribution. The naval-propulsion paper reports 99.24% accuracy on a public CODLAG dataset using a residual neural network with a fuzzy head, while the maritime paper pairs a neuro-fuzzy network with saliency maps and fuzzy rules for trustworthy decision-making. Evolve's reviews flag a consistent tension: strong headline metrics, but comparisons that may overstate superiority — a Table I tie with an SVM at 0.99 is read as undermining an 'outperforms SOTA' claim, and possible data leakage plus missing variance statistics weaken statistical support. Both papers carry low reproducibility-readiness ratings, giving a reproducibility mean of 20 across the two reviewed papers. Collaboration is compact: 3 distinct co-authors — Dionisis Kalogeropoulos, Georgia Sovatzidi, and Panagiotis G. Kalozoumis — yielding a collaboration score of 28. The radar focus score reads 0 despite a field distribution split evenly, 0.5 and 0.5, between two labels that describe essentially the same theme. Institution, country, ORCID, and citation-impact data are absent from this snapshot; the profile is name-resolved from public bibliographic records, so all scores should be read as provisional.
Collaboration pattern
Collaboration is tight and repeated: exactly 3 distinct co-authors — Dionisis Kalogeropoulos, Georgia Sovatzidi, and Panagiotis G. Kalozoumis — appear across the two reviewed arXiv preprints, yielding a collaboration score of 28. In both papers Iakovidis sits in the last author position without corresponding-author designation, a pattern consistent with a small, stable team in which a senior or supervisory role is implied while the corresponding role is held elsewhere. No cross-institution or larger consortium links are visible in this snapshot.
Research keywords
Highlights
Residual neuro-fuzzy pipeline reaches 99.24% on public naval-propulsion data
Saliency-plus-fuzzy-rule explanations for maritime decision-making
One explainability recipe applied across two maritime domains
Both reviewed papers reuse a single template — a neural predictor distilled into interpretable fuzzy decision trees — applied to naval propulsion maintenance and maritime trust. This consistency defines the methodological identity, though both papers carry low reproducibility-readiness ratings.
Timeline
- 2026Posts an arXiv preprint on explainable predictive condition-based maintenance of naval-propulsion systems, distilling a residual neural network with a fuzzy head into local fuzzy decision trees and reporting 99.24% accuracy on a public CODLAG dataset.
- 2026Posts a companion arXiv preprint on explainable neuro-fuzzy prediction for trustworthy decision-making in maritime settings, combining saliency maps with fuzzy-rule explanations.
- 2026Both preprints enter the Evolve review catalog with low reproducibility-readiness ratings, and review notes flag baseline-tie and possible-leakage concerns that temper the headline accuracy claims.
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.
- Bibliometric scores rest on a thin base: only 2 reviewed arXiv preprints, both from 2026, so all radar dimensions are provisional.
- Output, impact, and activity axes are unscored — no citation counts, h-index, or multi-year publication series are available in this snapshot.
- Institution, country, and ORCID are not provided; the profile is name-resolved from public bibliographic data and may merge or split same-name authors.
- Evolve verdicts quoted here are the platform's own review assessments of the two preprints, not independent peer review.