University of Otago ML researcher: federated learning submodelling and quantum granular-ball classification
2
papers here
2
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
23
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 publication-volume score is not yet computed. This axis would measure total research output and its consistency across the career record (1991–2026), but the underlying work counts are incomplete in the current bibliometric snapshot.
Impact—
A citation-impact score is not yet available. This axis would measure citation accumulation relative to the h-index (23) and i10-index (46), but per-paper citation and field-normalized metrics are missing from the current snapshot.
Focus0/100
The focus score is 0/100, the lowest possible reading, and the underlying data explain why: the two catalog fields split evenly (0.5 machine learning/quantum machine learning, 0.5 federated learning), and topic counts spread across five clusters of similar size — image retrieval and classification (30), neural network applications (26), image and video retrieval (25), neural networks (21), and video surveillance and tracking (20) — indicating broad topical dispersion rather than one narrow specialty.
Collaboration39/100
The collaboration score is 39/100, a moderate reading. The record lists 6 distinct recent co-authors, and the six recurring partners — Suzhen Yuan, Dehang Chen, Lifeng Shen, Shuyin Xia, Yasmeen Afzal, and Haibo Zhang — anchor a small, stable network. On both 2026 catalog papers Deng holds neither first nor corresponding position (middle on FedFIbOS, last on QGB-W$k$NN), consistent with a senior collaborator role rather than a lead-author pattern.
Activity100/100
The activity score is 100/100, the maximum. The record shows 61 works in the most recent three-year window, with annual counts climbing from 2 in 2023 to 10 in 2024, 14 in 2025, and 37 in 2026 — well above the long-run pattern that peaked at 12 in 2013. Sustained output since 1991 alongside an h-index of 23 and i10-index of 46 underpins the top reading.
Reproducibility38/100
The reproducibility score is 38/100, a mean over 2 reviewed papers. FedFIbOS rates medium: Evolve notes its optimality rests on strong unverified assumptions and that large gains are reported without seeds, variance, or code. QGB-W$k$NN rates low: classical-only simulation and anomalous noise results limit rigor. Both arXiv preprints would need released code and fuller experimental detail to raise this axis.
Research summary
Jeremiah D. Deng is a machine learning researcher at the University of Otago, New Zealand, whose OpenAlex record spans more than three decades, from a first recorded work in 1991 to 37 works in 2026. The record carries an h-index of 23 and an i10-index of 46, with 61 works in the most recent three-year window and a maximum annual count of 37 in 2026 — the sharpest upswing in a career that earlier peaked at 12 works in 2013. Topic modelling concentrates on image retrieval and classification (30), advanced neural network applications (26), image and video retrieval (25), neural networks (21), and video surveillance and tracking (20), while the current portfolio splits evenly between federated learning and quantum machine learning (0.5 each). Evolve has reviewed two 2026 arXiv papers: FedFIbOS, on Fisher-importance submodelling for heterogeneous federated learning (middle author), and QGB-W$k$NN, on quantum granular-ball classification (last author). Both are preprints with guarded verdicts — strong unverified assumptions and missing seeds, variance, or code in one case; classical-only simulation and anomalous noise results in the other — yielding a reproducibility mean of 38/100. Collaboration is moderate (39/100) across 6 distinct recent co-authors.
Collaboration pattern
Collaboration reads as a small, stable network: 6 distinct recent co-authors, with Suzhen Yuan, Dehang Chen, Lifeng Shen, Shuyin Xia, Yasmeen Afzal, and Haibo Zhang recurring most often. The 39/100 collaboration score reflects limited breadth rather than isolation. On both 2026 catalog papers Deng is neither first nor corresponding author — middle on FedFIbOS and last on QGB-W$k$NN — a pattern consistent with a senior contributor embedded in others' projects as well as his own.
Research keywords
Highlights
Fisher-importance submodelling for heterogeneous federated learning
Quantum granular-ball learning for robust classification
Sharpest output upswing in a three-decade record
Annual works rise from 2 in 2023 to 10 in 2024, 14 in 2025, and 37 in 2026, with 61 works across the last three years — the strongest sustained stretch since the 2013 peak of 12 works.
Timeline
- 1991Earliest recorded publication year in the OpenAlex record, with 1 work.
- 2000An early-career high of 5 recorded works.
- 2013Peak of the earlier career phase, with 12 recorded works.
- 2020Output rebounds to 8 works after the 2018–2019 dip (4 and 3 works).
- 2024Output climbs to 10 works, opening the current upswing.
- 202637 works recorded year-to-date; both Evolve-reviewed arXiv papers appear this year — FedFIbOS (federated learning) and QGB-W$k$NN (quantum granular-ball learning).
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 the OpenAlex record only (identity: openalex); output and impact axes lack the underlying data and are not scored.
- The reproducibility mean of 38 rests on only 2 Evolve-reviewed arXiv preprints, both from 2026.
- The 2026 annual count (37 works) covers a year still in progress at snapshot time and may shift.
- Topic labels and co-author names are name-resolved from public bibliographic data; no biographical details are included.