A Hybrid Modeling Framework for Crop Prediction Tasks via Dynamic Parameter Calibration and Multi-Task Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Solow, William, Pesantez-Cabrera, Paola, Keller, Markus, Khot, Lav, Saisubramanian, Sandhya, Fern, Alan
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917508778295296
author Solow, William
Pesantez-Cabrera, Paola
Keller, Markus
Khot, Lav
Saisubramanian, Sandhya
Fern, Alan
author_facet Solow, William
Pesantez-Cabrera, Paola
Keller, Markus
Khot, Lav
Saisubramanian, Sandhya
Fern, Alan
contents Accurate prediction of crop states (e.g., phenology stages and cold hardiness) is essential for timely farm management decisions such as irrigation, fertilization, and canopy management to optimize crop yield and quality. While traditional biophysical models can be used for season-long predictions, they lack the precision required for site-specific management. Deep learning methods are a compelling alternative, but can produce biologically unrealistic predictions and require large-scale data. We propose a \emph{hybrid modeling} approach that uses a neural network to parameterize a differentiable biophysical model and leverages multi-task learning for efficient data sharing across crop cultivars in data limited settings. By predicting the \emph{parameters} of the biophysical model, our approach improves the prediction accuracy while preserving biological realism. Empirical evaluation using real-world and synthetic datasets demonstrates that our method improves prediction accuracy by 60\% for phenology and 40\% for cold hardiness compared to deployed biophysical models.
format Preprint
id arxiv_https___arxiv_org_abs_2603_15411
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle A Hybrid Modeling Framework for Crop Prediction Tasks via Dynamic Parameter Calibration and Multi-Task Learning
Solow, William
Pesantez-Cabrera, Paola
Keller, Markus
Khot, Lav
Saisubramanian, Sandhya
Fern, Alan
Artificial Intelligence
Machine Learning
Accurate prediction of crop states (e.g., phenology stages and cold hardiness) is essential for timely farm management decisions such as irrigation, fertilization, and canopy management to optimize crop yield and quality. While traditional biophysical models can be used for season-long predictions, they lack the precision required for site-specific management. Deep learning methods are a compelling alternative, but can produce biologically unrealistic predictions and require large-scale data. We propose a \emph{hybrid modeling} approach that uses a neural network to parameterize a differentiable biophysical model and leverages multi-task learning for efficient data sharing across crop cultivars in data limited settings. By predicting the \emph{parameters} of the biophysical model, our approach improves the prediction accuracy while preserving biological realism. Empirical evaluation using real-world and synthetic datasets demonstrates that our method improves prediction accuracy by 60\% for phenology and 40\% for cold hardiness compared to deployed biophysical models.
title A Hybrid Modeling Framework for Crop Prediction Tasks via Dynamic Parameter Calibration and Multi-Task Learning
topic Artificial Intelligence
Machine Learning
url https://arxiv.org/abs/2603.15411