Yangdi Jiang

Nanyang Technological University

Unclaimed · AI-structured profileOpenAlex

Nanyang Technological University researcher: inference theory for manifold optimization and local differential privacy

papers here

reviewed by Evolve

first-author

corresponding

—

works · OpenAlex

—

h-index · OpenAlex

Bibliometric figures come from OpenAlex (2026-10-15) 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 data is available in this snapshot: publication counts and productivity series are missing, so this axis — which would measure the volume and cadence of scholarly output over time — remains unscored and is not estimated here.

Impact—

Impact data is unavailable: the snapshot carries no citation counts, h-index, or influence measures, so this axis — which would track how widely the work is cited and taken up by other researchers — is left unscored.

Focus0/100

Evolve's focus axis scores 0 for Yangdi Jiang, computed from a field distribution split exactly 50/50 between two clusters: 'mathematical statistics / differential privacy / Riemannian optimization' and 'mathematical statistics (nonparametric inference on Riemannian manifolds)'. The two lines are methodologically adjacent — both center on inference under geometric constraints — but the even split signals no dominant cluster yet, a pattern typical of an early two-paper snapshot rather than genuine diffusion across unrelated disciplines.

Collaboration36/100

The collaboration dimension scores 36, derived from 5 distinct co-authors in the snapshot. Both reviewed arXiv papers recur with the same two partners, Xiaotian Chang and Qirui Hu, suggesting a compact working group at Nanyang Technological University. On each paper Jiang sits in a middle author position and is not the corresponding author, a pattern consistent with a contributing specialist role inside a small team rather than an independent, correspondence-leading one.

Activity—

Activity data is missing: without indexed publication dates beyond the two 2026 preprints, this axis — which would measure recent publishing cadence and continued engagement with the field — cannot be scored and is not estimated.

Reproducibility55/100

Reproducibility scores 55, the mean readiness rating across exactly 2 reviewed papers, each rated medium. Evolve's verdicts temper the score with concrete reservations: the local-privacy paper's coverage leans on per-design safeguard tuning, and the tangent-field paper shows fragile small-sample coverage with no baseline comparison. Theoretical depth is consistently praised, but neither preprint yet demonstrates external replication or benchmarked baselines that would lift readiness above medium.

Research summary

Yangdi Jiang is a researcher at Nanyang Technological University whose reviewed work sits at the intersection of mathematical statistics, differential privacy, and optimization on Riemannian manifolds. Evolve's catalog contains two 2026 arXiv preprints, both with Jiang in a middle author position and neither as corresponding author. The first, 'Locally Private Inference for Riemannian Stochastic Optimization', claims a central limit theorem and transcript-based Wald inference for manifold-valued minimizers under local differential privacy; Evolve judges it rigorous, while noting that coverage leans on per-design safeguard tuning. The second, 'Riemannian Simultaneous Inference for Tangent Vector Field Regression', develops Gumbel-calibrated simultaneous confidence tubes for tangent field regression via kernel estimation and parallel transport; Evolve calls the theory strong but flags fragile small-sample coverage and the absence of baseline comparisons. Radar indicators computed from bibliometric data score collaboration at 36 (5 distinct co-authors, with Xiaotian Chang and Qirui Hu recurring), reproducibility at 55 (the mean over the two reviewed papers, both rated medium), and focus at 0, reflecting an even 50/50 split between two adjacent field clusters. Output, impact, and activity metrics are not yet available in the snapshot and are left unscored.

Collaboration pattern

The snapshot records 5 distinct co-authors and a collaboration radar score of 36. Both reviewed papers recur with the same partners, Xiaotian Chang and Qirui Hu, and place Jiang in a middle author position without corresponding-author status — a compact, repeat-partner pattern at Nanyang Technological University that suggests a specialist contributor within a small statistics team rather than a lab-leading correspondent. Broader network data is not yet indexed.

Research keywords

local differential privacyRiemannian stochastic optimizationPolyak-Ruppert averagingasymptotic inferenceHessian estimationRiemannian manifoldstangent vector fieldskernel regressionparallel transportsimultaneous confidence bandsGumbel limit

Highlights

CLT and Wald inference under local differential privacy on manifolds
The 2026 arXiv paper 'Locally Private Inference for Riemannian Stochastic Optimization' claims a central limit theorem and transcript-based Wald inference for manifold-valued minimizers under local DP. Evolve finds the arguments rigorous, while cautioning that coverage guarantees lean on per-design safeguard tuning.
Gumbel-calibrated simultaneous confidence tubes for tangent vector fields
'Riemannian Simultaneous Inference for Tangent Vector Field Regression' (2026, arXiv) constructs rigorous simultaneous tubes for manifold tangent field regression using kernel regression and parallel transport. Evolve rates the theory strong but notes fragile small-sample coverage and no baseline comparison.
A two-paper inference-first agenda, both rated medium reproducibility
Both reviewed 2026 preprints share one agenda — asymptotic guarantees under Riemannian-geometric constraints — and both carry medium reproducibility readiness, yielding the reproducibility mean of 55 across 2 papers recorded in the radar.

Timeline

2026Posts 'Riemannian Simultaneous Inference for Tangent Vector Field Regression' to arXiv in a middle author position.
2026Posts 'Locally Private Inference for Riemannian Stochastic Optimization' to arXiv in a middle author position.
2026Both preprints enter the Evolve catalog with medium reproducibility-readiness ratings; the radar records 5 distinct co-authors and leaves output, impact, and activity unscored.

Papers in the Evolve catalog

Frequent co-authors

Xiaotian Chang×2 Qirui Hu×2 Xiaotian Chang×1 Yangdi Jiang×1 Qirui Hu×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 and radar scores come from an OpenAlex-based snapshot; citation counts, h-index, and publication totals are absent, leaving output, impact, and activity unscored.
Profile is name-resolved from public bibliographic data; ORCID and country are not provided and 'Yangdi Jiang' is not a rare name, so identity confidence is limited.
Only two papers, both 2026 arXiv preprints, have been reviewed by Evolve; venue acceptance and earlier work fall outside this snapshot.
Focus and collaboration scores reflect this narrow snapshot (two field clusters, 5 distinct co-authors) and may shift as more publications are indexed.