Try:droughtFusariumnitrogen fertilizerArabidopsisseed germination

“model performance”

9,388 matches

Claim types

Direction of the findings

plus 2,296 non-evaluative

Over time

1960 ’70 ’80 ’90 2000 ’10 ’20

Refine

More filters

Method / resourceobservedoriginal resultpositivemeasured

Claim · model-written summary

Model performance indicators were satisfactory.

Taxon: potatoSolanum tuberosumt0
Outcome: model performancee0
Qualifier: satisfactorye0

Not reportedComparatorValue

Verbatim

e0
“The indicators of model performance also showed that the model performance was satisfactory.”
✓ located verbatim in the source · abstract, characters 952–1044

t0
“potato”
✓ located verbatim in the source · title, characters 32–38

Title

Calibration of WOFOST model for potato in India

Vikas Dua; PM Govindakrishnan; B. P. Singh Potato Journal/Journal of the Indian Potato Association 2014

Other claims from this paper (6)

Method / resource Simulated potato emergence differed from measurements by 0 to 2 days.
Method / resource Simulated potato tuber growth initiation differed from measurements by 1 to 8 days.
Method / resource Simulated potato physiological maturity differed from measurements by 5 to 12 days.
Method / resource Final dry matter yield variation ranged from 5.6 to 7.5%, within acceptable limits.
Recommendation The authors concluded WOFOST can be used for potato under Indian conditions.
Recommendation The authors advised validation under varied conditions before wide use.

Method / resourceobservedoriginal resultpositivemeasured

Claim · model-written summary

Reduced model performance was very close to full model performance.

Taxon: sugarcanee0
Outcome: performancee1
Direction: very closee1
Comparator: full modele1

Verbatim

e0
“sugarcane ( Saccharum officinarum L.)”
✓ located verbatim in the source · abstract, characters 93–130

e1
“its performance is very close to that of the full model”
✓ located verbatim in the source · abstract, characters 1471–1526

Title

Can statistical learning models make early selection among sugarcane families easier and still efficient?

Édimo Fernando Alves Moreira; Marcio Henrique Pereira Barbosa; Luiz Alexandre Peternelli Crop Science 2020. doi:10.1002/csc2.20334

Other claims from this paper (4)

Method / resource All classifiers had low apparent error rates in full and reduced models.
Method / resource All classifiers had high true positive rates in full and reduced models.
Method / resource SVM had the best classifier performance.
Recommendation Authors prefer the reduced model because its operation is more straightforward.

Method / resourceobservedoriginal resultmeasured

Claim · model-written summary

Model performance was sensitive to parametrization method.

Outcome: Model performancee0
Comparator: applied parametrization methode0

Verbatim

e0
“Model performance was sensitive to the applied parametrization method”
✓ located verbatim in the source · abstract, characters 1198–1267

Title

Comparing Apple and Pear Phenology and Model Performance: What Seven Decades of Observations Reveal

Bianca Drepper; Anne Gobin; Serge Remy; et al. Agronomy 2020. doi:10.3390/agronomy10010073

Other claims from this paper (8)

Association Apple flowering was 9.5 days earlier after warm dormancy periods.
Association Pear flowering was 11.5 days earlier after warm dormancy periods.
Association The bloom date and temperature relationship was nonlinear.
Observation After warm chilling, models predicted apple dormancy break 5.0 days later.
Observation After warm chilling, models predicted pear dormancy break 10.6 days later.
Method / resource PLSR gave mixed signals about flowering onset after warm chilling.
Method / resource A dynamic chill model with growing degree days had lower flowering prediction errors than a continuous chill-forcing model.
Method / resource Application of both models outside prior temperature ranges showed limitations.

Method / resourceobservedoriginal resultpositivesignificant (as stated)measured

Claim · model-written summary

Hyperparameter tuning improved model performance.

Taxon: Sugarcanet0
Outcome: model performancee0
Direction: enhancede0

Not reportedComparator

Verbatim

e0
“Hyperparameter tuning significantly enhanced model performance”
✓ located verbatim in the source · abstract, characters 612–674

t0
“Sugarcane”
✓ located verbatim in the source · title, characters 50–59

Title

Advanced Machine Learning Approaches for Accurate Sugarcane Yield Prediction with Ensemble Techniques

Md. Sorowar Mahabub Rabby; Tajniya Nowshin Ome; Mst. Maliha Mobassira; et al. 2025 International Conference on Electrical, Computer and Communication Engineering (ECCE) 2025. doi:10.1109/ecce64574.2025.11013297

Other claims from this paper (5)

Method / resource Gradient Boosting achieved an R2 score of 99.38%.
Method / resource Random Forest achieved an R2 score of 99.24%.
Method / resource Stacking Ensemble outperformed individual models with an R2 score of 99.24%.
Method / resource Stacking Ensemble had the lowest MSE, 7.20.
Method / resource Feature importance analysis identified key drivers of yield.

Method / resourceobservedoriginal resultno effectnot significant (as stated)measured

Claim · model-written summary

Compact-model performance was not significantly different from full-model performance at α = 0.05.

Taxon: grapevineVitist0
Outcome: performancee0
Direction: not significantly differente0
Value: α = 0.05e0
Comparator: full modele0

Verbatim

e0
“with performance not significantly different from the full model at α = 0.05”
✓ located verbatim in the source · abstract, characters 1093–1169

t0
“grapevine leaf water potential”
✓ located verbatim in the source · abstract, characters 525–555

Title

A machine learning framework for California vineyard water status monitoring using sUAS Imagery and short-term meteorological data

Rui Gao; Maria Mar Alsina; Alfonso F. Torres-Rua; et al. Irrigation Science 2026. doi:10.1007/s00271-026-01102-8

Other claims from this paper (6)

Method / resource The full XGB model achieved R² of 0.71 for leaf water potential on the independent test data.
Method / resource The full XGB model had RMSE of 0.14 MPa.
Method / resource The full XGB model had bias of − 0.06 MPa.
Method / resource The compact XGB model achieved R² of 0.63.
Method / resource The compact XGB model had RMSE of 0.16 MPa.
Method / resource The compact XGB model had bias of − 0.06 MPa.

Method / resourceobservedoriginal resultpositivemeasured

Claim · model-written summary

PReLU enhanced model performance.

Outcome: model performancee0
Direction: enhancese0

Not reportedTaxonComparator

Verbatim

e0
“PReLU reduces overfitting and enhances model performance”
✓ located verbatim in the source · abstract, characters 938–994

Title

Classification and Detection of Rice Diseases Using a 3-Stage CNN Architecture with Transfer Learning Approach

Munmi Gogoi; Vikash Kumar; Shahin Begum; et al. Agriculture 2023. doi:10.3390/agriculture13081505

Other claims from this paper (4)

Method / resource The approach reduced the training data required while retaining high accuracy.
Method / resource Progressive resizing improved feature learning.
Method / resource PReLU reduced overfitting.
Method / resource The CNN achieved 94% accuracy on rice leaf images with ten-fold cross-validation, significantly higher than other methods.

Method / resourceobservedoriginal resultmeasured

Claim · model-written summary

An ablation study found architectural choices affected model performance.

Taxon: MaizeZea mayst0
Outcome: model performancee0
Comparator: architectural choicee0

Verbatim

e0
“through an ablation study of architectural choice on model performance, the value of certain architectural choices on model performance”
✓ located verbatim in the source · abstract, characters 913–1048

t0
“Maize Cob”
✓ located verbatim in the source · title, characters 39–48

Title

MaizeNet: High-Performance Image-Based Maize Cob Detection using Lightweight CNNs

Nidhi Mishra; Banti Lalnunthari Springer Link (Chiba Institute of Technology) 2025. doi:10.1051/shsconf/202521601063/pdf

Other claims from this paper (4)

Method / resource MaizeNet detected maize cobs with 91.4% mean average precision.
Method / resource MaizeNet processed 25 frames per second on standard mobile hardware.
Method / resource Comparative evaluations found MaizeNet outperformed other deep learning models.
Method / resource The authors report MaizeNet robustness under diverse field conditions and scalability to larger datasets.

Method / resourceobservedoriginal resultno effectmeasured

Claim · model-written summary

Greater model complexity did not notably improve detection performance.

Taxon: Siam Weedt0
Outcome: model performancee0
Direction: did not notably enhancee0
Comparator: Increased model complexitye0

Verbatim

e0
“Increased model complexity did not notably enhance model performance”
✓ located verbatim in the source · abstract, characters 1531–1599

t0
“Siam Weed”
✓ located verbatim in the source · title, characters 31–40

Title

Detection of Invasive Species (Siam Weed) Using Drone-Based Imaging and YOLO Deep Learning Model

Deepak Gautam; Zulfadli Mawardi; Louis Elliott; et al. Remote Sensing 2025. doi:10.3390/rs17010120

Other claims from this paper (5)

Method / resource YOLOv5 achieved an F1 score above 0.85 for Siam weed detection.
Method / resource Improved image exposure raised the detection F1 score above 0.95.
Method / resource Approximately 1000 image tiles trained a reliable detection model; additional tiles gave marginal improvement.
Method / resource Foliage and bark under high solar illumination often caused false positives.
Method / resource Low exposure images considerably reduced false positives.

Method / resourceobservedoriginal resultsignificant (as stated)measured

Claim · model-written summary

Processed sample preparation significantly improved model performance.

Outcome: model performancee0
Direction: improvede0

Not reportedComparatorValue

Verbatim

e0
“processed sample preparation significantly improved model performance”
✓ located verbatim in the source · abstract, characters 1408–1477

Title

Near‐Infrared Spectroscopy Prediction of Dry Matter and Starch Content in Cassava Using Optimized Calibration Models

Paulo Henrique Ramos Guimarães; Massaine Bandeira e Sousa; Marcos de Souza Campos; et al. Journal of Food Science 2025. doi:10.1111/1750-3841.70704

Other claims from this paper (8)

Method / resource PLS delivered highest predictive accuracy across traits and devices.
Method / resource KNN slightly outperformed PLS for DMCg with the benchtop device.
Method / resource XGB was comparable to PLS for StC on the benchtop device.
Method / resource XGB was comparable to PLS for DMCo on the portable device.
Method / resource Processed samples produced higher model accuracy than fresh samples.
Method / resource Portable NIR performed better with processed samples.
Method / resource Portable NIR surpassed benchtop DMCg validation accuracy.
Method / resource Portable NIR surpassed benchtop StC validation accuracy.

Method / resourceobservedoriginal resultpositivemeasured

Claim · model-written summary

Adding crown length improved sapwood model performance.

Outcome: model performancee0
Direction: improvede0
Comparator: sapwood equationse0

Not reportedTaxon

Verbatim

e0
“adding crown length to the sapwood equations improved model performance.”
✓ located verbatim in the source · abstract, characters 330–402

Title

Leaf area prediction models for Tsuga canadensis in Maine

Laura S Kenefic; Robert S Seymour Canadian Journal of Forest Research 1999. doi:10.1139/x99-134

Other claims from this paper (5)

Method / resource Sapwood area was an effective hemlock leaf-area surrogate.
Method / resource The best sapwood equation had upper-range prediction bias.
Method / resource Sapwood area at crown base predicted leaf area less well than at breast height.
Method / resource Crown length or volume alone produced the least effective models.
Method / resource Stem cross-sectional area or basal-area models performed comparably to the best sapwood model without bias.

Lexical search (BM25). All words must match; if nothing does, any word. Organism names also match every spelling of the same organism. No semantic search yet.

“model performance”

Claim types

Direction of the findings

Over time

Claim type

Research area

Polarity

Statistics