Markus Keller

Washington State University · United States

Unclaimed · AI-structured profileORCID OpenAlex

Washington State University researcher bridging viticulture and plant physiology with machine-learning crop modeling

papers here

reviewed by Evolve

first-author

corresponding

—

works · OpenAlex

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 scored: no per-year publication vector was provided for this axis. It would measure total research output and the productivity trajectory; the snapshot lists 33 works in the last three years but yields no output score.

Impact—

Not scored: the snapshot lacks the citation-per-paper vector this axis needs. It would measure field-normalized citation influence; h-index 42 and i10-index 94 are recorded as raw values but were not converted into an impact score.

Focus100/100

Focus scores 100/100 on topic concentration: OpenAlex clusters 151 works in horticultural and viticultural research, 98 in plant physiology and cultivation, and 71 in berry genetics and cultivation, with smaller strands of 25 in irrigation and water management and 14 in plant water relations. One catalog paper sits in machine learning for agriculture — the sole entry in the field distribution — a narrow, newly grafted branch on an otherwise concentrated viticultural core.

Collaboration48/100

Collaboration scores 48/100: the recent window shows 10 distinct coauthors against 33 works in the last three years, a compact network. The frequent-coauthor list — Paola Pesántez-Cabrera, Alan Fern, William Solow, Lav Khot, Sandhya Saisubramanian and others — points to a Washington State University team mixing horticulture and computing. On both reviewed 2026 papers he is a middle author without corresponding status, and last author on the LSTM berry-temperature preprint.

Activity100/100

Activity scores 100/100. The OpenAlex record spans 1991 to 2026 with annual counts such as 8 in 2004, 16 in 2015, 13 in 2020, and 16 in 2025 — the record's peak — plus 13 logged in 2026 to date. The last three years alone contribute 33 works, confirming an unusually sustained and current publishing rhythm across a 35-year indexed career.

Reproducibility20/100

Reproducibility scores 20/100, computed from Evolve's assessments (reproMean 20 across 1 repro-scored paper). Of the two reviewed preprints, the multi-task cold-hardiness study is rated low — broad results but no statistical significance testing — while the cross-regional transfer-learning paper rates medium, with a flagged risk that text priors leak label information. No code or data availability evidence appears in the snapshot.

Research summary

Markus Keller, profiled at Washington State University, is a researcher whose OpenAlex record pairs a deep viticulture and horticulture core with a recent, visible move into machine-learning crop modeling. The bibliometric snapshot is strong on continuity: an h-index of 42 and i10-index of 94, with works logged every year from 1991 through 2026, peaking at 16 publications in both 2015 and 2025 and already 13 in 2026; the last three years alone contribute 33 works. Topic clustering is highly concentrated — 151 works in horticultural and viticultural research, 98 in plant physiology and cultivation, 71 in berry genetics and cultivation, plus 25 in irrigation and water management and 14 in plant water relations — yielding focus and activity scores of 100/100, against collaboration at 48 (10 distinct recent coauthors) and reproducibility at 20. The Evolve catalog adds three 2026 arXiv preprints where he appears as a middle or last author: a multi-task learning case study for grape cold hardiness and budbreak that outperforms single-task and Ferguson baselines (though without statistical significance testing), a transfer-learning framework for cross-regional LT50 prediction built on text and few-shot embeddings, and an LSTM-based grape berry temperature forecaster. Output and impact axes remain unscored for lack of a citation vector.

Collaboration pattern

Collaboration score 48/100 reflects a compact recent network: 10 distinct coauthors over the last three years against 33 works. The catalog papers place him in a WSU-centered team — Paola Pesántez-Cabrera, Alan Fern, Lav Khot, Sandhya Saisubramanian and others — spanning horticulture and computer science; on the reviewed arXiv papers he is a middle or last author, not corresponding, indicating contribution within larger group efforts rather than solo leadership.

Research keywords

viticulturegrape cold hardinessLT50 predictionbudbreak predictionmulti-task learningtime-series modelingrecurrent neural networkstransfer learningLSTM forecastingprecision viticultureirrigation and water managementplant water relations

Highlights

Wrapper multi-task learning lifts grape cold-hardiness and budbreak prediction
Evolve's review finds a wrapper multi-task learning approach beating single-task and Ferguson models for grape cold hardiness and budbreak on sparsely labeled time series, combining task embeddings with recurrent neural networks. Results are broad, but the study lacks statistical significance testing and repro-readiness is rated low.
First transfer-learning framework for cross-regional grapevine LT50
The paper proposes learning multimodal latent representations from text and few-shot embeddings to transfer grapevine LT50 cold-hardiness prediction across regions — per Evolve, the first such transfer-learning framework. Results are judged credible, but text priors may leak label information. Repro-readiness: medium.
LSTM with feed-forward attention for multi-step grape berry temperature forecasting
A 2026 arXiv preprint, with Keller as last author, applies an LSTM model with feed-forward attention to multi-step forecasting of grape berry temperature. Evolve has not yet reviewed it: no verdict, keywords, or repro-readiness score are on record.

Timeline

2026Co-authored (middle position) the multi-task learning case study on sparsely labeled time series for grape cold hardiness and budbreak prediction.
2026Middle author on the first transfer-learning framework for cross-regional grapevine LT50 cold-hardiness prediction via learned multimodal latent representations.
2026Last author on the LSTM-with-feed-forward-attention preprint forecasting grape berry temperature over multiple steps.
2026All three Evolve-catalog papers appeared as arXiv preprints this year, marking a visible machine-learning extension of a long viticultural publishing record.

Papers in the Evolve catalog

Frequent co-authors

Paola Pesántez-Cabrera×2 Alan Fern×2 William Solow×1 Lav Khot×1 Sandhya Saisubramanian×1 Srikanth Gorthi×1 L. G. Divyanth×1 Dattatray Bhalekar×1 Markus Keller×1 Aseem Saxena×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 figures come from an OpenAlex snapshot (identity: openalex); h-index 42 and i10-index 94 reflect that index only, not a full bibliographic audit.
Name resolution note: 'Markus Keller' also appears within the frequent-coauthor list itself, and horticulture-heavy topics sit beside recent ML preprints, so author-disambiguation noise cannot be ruled out; this profile is name-resolved from public bibliographic data.
The 2026 preprint 'Multi-Step Forecasting of Grape Berry Temperature based on LSTM Model with Feed-Forward Attention' has no Evolve review, keywords, or repro-readiness assessment yet.
Output and impact radar axes are unscored (insufficient underlying data), the 2026 counts are partial-year, and no per-paper citation vector was provided.