Multi task learning improves cold hardiness prediction for grapevines
Multi-Task Learning for Sparsely-Labeled Time Series: A Case Study on Cold-Hardiness Modeling
Machine Learning
Summary
Predicting the temperature at which grapevines get damaged by cold is important for farmers, but getting enough data is hard because measurements are rare. The authors use a type of AI called multi-task learning that learns from several grape varieties at once, improving predictions even with limited data. Their approach also works well for predicting when grape buds start to grow in spring by learning these related tasks together. This method beats existing models and can help farmers better decide when to protect their crops from frost.
What this means in practice
- •For agriculture technology teams: Develop predictive tools that estimate grapevine cold damage risk across multiple cultivars using limited field data.$Commercial implications: Enables commercial frost protection advisory services tailored to different grape varieties to reduce crop losses.
- •For orchard management teams: Use multi-task trained models to forecast both cold damage and budbreak timing to optimize frost protection and growth management.
Authors
Aseem Saxena, Paola Pesántez-Cabrera, Jonathan Magby, Markus Keller, Alan Fern
Abstract
We present a real-world case study of multi-task learning (MTL) for temporal process modeling from limited data with temporally sparse labels. Specifically, we investigate multi-task learning for the important agricultural problem of predicting grape cold hardiness, which is the temperature at which lethal freezing occurs. Cold hardiness changes in response to weather and is difficult to measure directly in the field. Thus, growers rely on predictions to decide when to apply costly frost mitigation measures. We apply recurrent neural networks (RNNs) for daily cold-hardiness prediction from time series weather data. A major challenge is that the cold hardiness response varies across plant cultivars and ground-truth data for each cultivar is temporally sparse and limited. To address this challenge, we investigate multi-task learning (MTL) approaches for combining data, where different tasks correspond to different cultivars. We develop a variety of MTL architectures and evaluate them in both MTL and transfer learning settings. Our results show significant differences between architectures and that certain architectures are able to consistently outperform single-task learning and state-of-the-art scientific models. Additionally, we show similar results for the qualitatively different, but related, task of budbreak prediction. Further, improved accuracy for budbreak and cold hardiness is achieved by a single MTL model that simultaneously learns both tasks.