Elena Raponi

Leiden University · Netherlands

Unclaimed · AI-structured profileORCID OpenAlex

Leiden University researcher in evolutionary optimization and robustness benchmarks for time-series virtual sensing.

papers here

reviewed by Evolve

first-author

corresponding

—

works · OpenAlex

h-index · OpenAlex

Bibliometric figures come from OpenAlex (2026-10-13) 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: this snapshot lacks the verified per-work output data (full publication list and venues) the axis needs. It would measure publication volume, consistency, and authorship roles over time; only per-year tallies and two reviewed papers are on record.

Impact—

Not yet scored: citation-level inputs (citations per paper, venue influence) are unavailable, so no impact value is estimated. The axis would measure uptake and downstream influence of the work; an h-index of 14 is on record but is not sufficient to render the axis.

Focus0/100

The radar assigns focus a score of 0, and the raw metrics support a picture of breadth over concentration: the field distribution splits evenly (0.5/0.5) between machine learning and robustness benchmarks for time-series and cyber-physical systems, while the five largest topic clusters span multi-objective optimization (47), metaheuristics (38), evolutionary algorithms (17), bandit algorithms (16), and topology optimization (13) — a method-spanning portfolio rather than one narrow lane.

Collaboration51/100

Collaboration scores 51, a mid-range value consistent with 12 distinct coauthors and a compact recurring core: Noah C. Puetz, Jens U. Brandt, Marc Hilbert, Alexander Windmann, Thomas Bäck, and Thomas Bartz-Beielstein appear repeatedly in the frequent-coauthor list. Both Evolve-reviewed papers position her as a middle author without corresponding role, matching a pattern of team-built benchmark papers within a small, stable group rather than large ad-hoc consortia.

Activity100/100

Activity tops the radar at 100. The yearly counts climb from 1 work in 2016 to 10 in 2022, 9 in 2023, 9 in 2024, 14 in 2025, and 16 in 2026 to date, with 39 works in the last three years — a sustained acceleration in output tempo that anchors this maximum score.

Reproducibility100/100

Reproducibility also scores the maximum 100, computed as a mean of 100 across the 2 Evolve-reviewed papers, both rated high reproducibility readiness. Evolve's verdicts cite solid multi-seed evidence for the bMASE/MuViS-DIR work and single-seed evidence for MuViS-C — a transparent reporting style that still invites follow-up significance testing.

Research summary

Elena Raponi is a researcher at Leiden University (NL) whose OpenAlex-resolved profile centers on optimization-driven machine learning: the largest topic clusters are advanced multi-objective optimization algorithms (47 works), metaheuristic optimization (38), evolutionary algorithms (17), bandit algorithms (16), and topology optimization in engineering (13). In the Evolve catalog she appears as a middle author, non-corresponding, on two 2026 arXiv preprints that build evaluation infrastructure for deep learning on time series: one exposes blind spots in deep imbalanced regression evaluation through the bMASE metric and MuViS-DIR benchmark; the other, MuViS-C, is a multi-domain robustness benchmark showing all virtual sensors degrade under sensor failures, with tree-based models the most robust. Her field distribution splits evenly (0.5/0.5) between machine learning and robustness benchmarking for time-series and cyber-physical systems. Output is accelerating: 39 works in the last three years, rising from 9 (2024) to 14 (2025) and 16 (2026 to date), after earlier years in single digits. Recorded citation indicators are an h-index of 14 and an i10-index of 17. Both reviewed papers reach the maximum reproducibility mean (100 across 2 papers).

Collaboration pattern

The collaboration score of 51 reflects a moderately sized, stable network: 12 distinct coauthors, with Noah C. Puetz, Jens U. Brandt, Marc Hilbert, Alexander Windmann, Thomas Bäck, and Thomas Bartz-Beielstein recurring most often in the supplied list. Both reviewed 2026 papers are team efforts with middle authorship and no corresponding role, suggesting she contributes as a core member of a compact benchmark-building group rather than leading large distributed consortia.

Research keywords

Multi-objective optimizationMetaheuristicsEvolutionary algorithmsBandit algorithmsTopology optimizationDeep imbalanced regressionEvaluation methodologyBenchmarkingVirtual sensingTime-series extrinsic regressionSensor failure robustnessDistribution shift

Highlights

Exposing evaluation blind spots in deep imbalanced regression
The 2026 arXiv preprint proposes the bMASE metric and the MuViS-DIR benchmark to close evaluation gaps in deep imbalanced regression, touching time-series extrinsic regression and virtual sensing; Evolve finds the multi-seed evidence solid but notes a single backbone and no significance tests weaken model rankings.
Multi-domain benchmark shows virtual sensors fail under corruption
MuViS-C benchmarks robustness under sensor failures across domains; the verdict is that all virtual sensors degrade under corruption and tree-based models are the most robust — a credible benchmark, though the evidence rests on a single seed.
A reproducibility-first 2026 benchmarking program
Both 2026 preprints reach the maximum reproducibility mean (100 across 2 papers, high readiness), forming a paired program of evaluation infrastructure for deep imbalanced regression and reliable virtual sensing.

Timeline

2016First year in the OpenAlex publication profile, with 1 recorded work; annual output stays low through 2020, peaking at 4 works in 2019.
2021Output climbs to 6 works, a clear step up from the single-digit start of the decade.
2022A step change to 10 works — the highest annual total recorded before 2025.
2025The annual count reaches 14 works, continuing the rise from 9 in 2024.
2026Year-to-date count of 16 works; two Evolve-reviewed arXiv preprints appear — the bMASE/MuViS-DIR imbalanced-regression evaluation study and the MuViS-C virtual-sensing robustness benchmark.

Papers in the Evolve catalog

Frequent co-authors

Noah C. Puetz×1 Jens U. Brandt×1 Marc Hilbert×1 Elena Raponi×1 Thomas Bäck×1 Thomas Bartz-Beielstein×1 Jens U. Brandt×1 Noah C. Puetz×1 Alexander Windmann×1 Marc Hilbert×1 Thomas B\"ack×1 Thomas Bartz-Beielstein×1

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 indicators (h-index 14, i10-index 17, per-year counts, topics) come from an OpenAlex-resolved profile (identity: openalex); name resolution should be confirmed against ORCID 0000-0001-6841-7409.
The output and impact radar axes are unscored because the underlying per-work data is insufficient in this snapshot.
Only the two 2026 arXiv preprints carry full Evolve reviews; the rest of the publication record is not individually assessed.
The supplied frequent-coauthor list contained duplicated entries and one self-record (Elena Raponi); duplicates and the self-entry were removed before counting.