Penn State statistician linking sufficient dimension reduction, sparse high-dimensional inference, and learning theory
2
papers here
1
reviewed by Evolve
0
first-author
0
corresponding
—
works · OpenAlex
21
h-index · OpenAlex
Bibliometric figures come from OpenAlex (2026-10-17) 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: the snapshot lacks the completeness signals this axis needs. It would measure total publication volume and productivity trajectory, which here can only be glimpsed indirectly through per-year OpenAlex counts, so no score or estimate is offered.
Impact—
Not yet scored: citation and influence data are absent from this snapshot. The axis would measure citation-based reach — total citations, citing works, and uptake of methods — none of which can be inferred from the available bibliographic record.
Focus100/100
Extremely concentrated: all five leading OpenAlex topics are statistical — Statistical Methods and Inference (84), Advanced Statistical Methods and Models (35), Statistical Methods and Applications (33), Sparse and Compressive Sensing Techniques (30), and Random Matrices and Applications (21) — and the field distribution lists a single field, statistics/machine learning (sufficient dimension reduction and generative modeling). The 2026 reviewed papers extend this core into generative modeling and reinforcement learning theory without leaving the methodological lane.
Collaboration32/100
Modest by the numbers, scoring 32 here: the recent three-year window shows 65 works but only 4 distinct coauthors, with recurring partners Zihang Liang, Haochen Zhang, Wenxi Tan, and Bing Li. Both reviewed 2026 papers place Xue as last author rather than corresponding author, consistent with a senior-collaborator pattern inside small, stable teams rather than broad external networks.
Activity100/100
Sustained and accelerating, earning a top activity score: yearly OpenAlex counts rise from 13 works in 2019 and 17 in 2021 to 20 in 2024, 21 in 2025, and 24 in 2026, giving 65 works in the recent three-year window — the highest sustained pace anywhere in the 2007–2026 record.
Reproducibility55/100
Intermediate (mean 55) with thin evidence: the score averages a single Evolve-reviewed paper (reproPapers: 1). Both reviewed 2026 works are rated medium reproducibility readiness — the engression paper's experiments presuppose the assumed low-dimensional structure, while Fed-LSVI offers rigorous proofs but no experiments — so the aggregate figure should be read cautiously.
Research summary
Lingzhou Xue is a statistician at Pennsylvania State University whose OpenAlex-resolved record centers on statistical methodology, with an h-index of 21 and an i10-index of 39. Yearly output has climbed steadily: 13 works in 2019, 17 in 2021, 20 in 2024, 21 in 2025, and 24 in 2026, totaling 65 works over the recent three-year window. Topic data keeps the portfolio squarely in methodology — Statistical Methods and Inference (84), Advanced Statistical Methods and Models (35), Statistical Methods and Applications (33), Sparse and Compressive Sensing Techniques (30), and Random Matrices and Applications (21). Evolve's reviewed catalog adds two 2026 arXiv papers with Xue as last author. Belted Engression embeds a sufficient dimension reduction bottleneck in Engression for generative distributional regression, with theory delivering faster rates and smaller networks, though experiments presuppose the assumed low-dimensional structure. Fed-LSVI gives the first provably efficient federated online reinforcement learning for linear MDPs, with Õ(√(Md³H⁴T)) regret and logarithmic communication cost, supported by rigorous proofs but no experiments. Frequent coauthors include Zihang Liang, Haochen Zhang, Wenxi Tan, and Bing Li. Output and impact remain unscored because citation-volume data is missing from this snapshot.
Collaboration pattern
The recent record shows a small, stable team: 65 works over three years with only 4 distinct coauthors. Recurring partners are Zihang Liang, Haochen Zhang, Wenxi Tan, and Bing Li. In both Evolve-reviewed 2026 papers Xue takes the last-author position without being corresponding author, consistent with supervising junior first authors on theory-heavy projects. The low collaboration score (32) reflects the narrow coauthor set rather than an absence of collaboration.
Research keywords
Highlights
Sufficient dimension reduction meets engression
First provably efficient federated online RL for linear MDPs
Output accelerating to a record pace
OpenAlex yearly counts climb from 13 works in 2019 and 17 in 2021 to 20 in 2024, 21 in 2025, and 24 in 2026 — 65 works in the recent three-year window, the fastest sustained stretch in the 2007–2026 record.
Timeline
- 2007Earliest recorded works appear in the OpenAlex yearly counts (3 publications), with only sparse output in the years before 2012.
- 2019Output steps up to 13 works, starting a sustained rise beyond the 10 works recorded in 2017.
- 202117 works recorded, up from 10 in 2020 — nearly double in a single year.
- 202420 works, opening a three-year surge (21 in 2025 and 24 in 2026) that totals 65 works.
- 2026Two Evolve-reviewed arXiv papers list Xue as last author: Belted Engression (sufficient dimension reduction for generative distributional regression) and Fed-LSVI (provably efficient federated reinforcement learning with logarithmic communication cost).
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 figures (h-index 21, i10-index 39, yearly counts) come from a single OpenAlex snapshot.
- Output and impact dimensions are unscored because the underlying citation-volume data is missing from this snapshot.
- The reproducibility score (mean 55) averages only one reviewed paper, so it is statistically thin.
- Identity is name-resolved from public bibliographic data (OpenAlex; ORCID 0000-0002-8252-0637); only the two listed 2026 arXiv papers have been individually reviewed by Evolve.