Genomic Prediction of Wheat Grain Yield Using Machine Learning
Problem
Breeders need to predict wheat grain yield from genetic markers, but the best algorithms are unclear.
Why it matters
Wheat grain yield prediction is critical for breeding staple crops that feed the world.
What the AI did
Compared machine learning and deep learning algorithms against classical Bayesian approaches across feature selection methods for predicting wheat grain yield on three datasets.
Results
Tree-based ML (random forests, gradient boosting) performed best; algorithm choice mattered more than feature selection; Bayesian methods prone to fitting problems
Impact
Findings guide algorithm choice for genomic prediction of wheat grain yield.
Record
| Venue | Agriculture |
|---|---|
| Published | 2022 |
| Work type | article |
| Authors | Manisha Sanjay Sirsat, Paula Rodrigues Oblessuc, Ricardo S. Ramiro |
| Citations | 35 (OpenAlex, at import time) |
| References | 54 |
| Model or method | Random forests, gradient boosting, deep learning, Bayesian methods |
| Organization | not stated in source |
| Primary source | https://doi.org/10.3390/agriculture12091406 |
| Open access | gold |
| Keywords | genomic prediction, wheat, machine learning, random forests, feature selection |
Importance rating
significant — thousands to millions affected
Abstract states genomic prediction is critical for grain yield of staple crops like wheat, crucial to feeding the world; guides breeding of a global staple, affecting millions via food supply.