Dynamic Prediction of Chilo suppressalis Occurrence in Rice Based on Deep Learning
Open Access
Abstract
(1) Background: The striped rice stem borer (SRSB), Chilo suppressalis, has severely diminished the yield and quality of rice in China. A timely and accurate prediction of the rice pest population can facilitate the designation of a pest control strategy. (2) Methods: In this study, we applied multiple linear regression (MLR), gradient boosting decision tree (GBDT), and deep auto-regressive (DeepAR) models in the dynamic prediction of the SRSB population occurrence during the crop season from 2000 to 2020 in Hunan province, China, by using weather factors and time series of related pests. (3) Results: This research demonstrated the potential of the deep learning method used in integrated pest management through the qualitative and quantitative evaluation of a reasonable validating dataset (the average coefficient of determination for the DeepAR, GBDT, and MLR models were 0.952, 0.500, and 0.166, respectively). (4) Conclusions: The DeepAR model with integrated ground-based meteorological variables, time series of related pests, and time features achieved the most accurate dynamic forecasting of the population occurrence quantity of SRSB as compared with MLR and GBDT.
Keywords
Funding Information
- the National Science Fund Project of China (31772157)
This publication has 33 references indexed in Scilit:
- Deep Learning: Methods and ApplicationsFoundations and Trends® in Signal Processing, 2014
- Representation Learning: A Review and New PerspectivesIEEE Transactions on Pattern Analysis and Machine Intelligence, 2013
- Insect-Resistant Genetically Modified Rice in China: From Research to CommercializationAnnual Review of Entomology, 2011
- Stepwise Multiple Regression Method to Forecast Fish LandingProcedia - Social and Behavioral Sciences, 2010
- Learning Deep Architectures for AIFoundations and Trends in Machine Learning, 2009
- Assessments of crop losses in rice ecosystems due to stem borer damage (Lepidoptera: Pyralidae)Crop Protection, 2006
- How to analyze long‐term insect population dynamics under climate change: 50‐year data of three insect pests in paddy fieldsPopulation Ecology, 2005
- Greedy function approximation: A gradient boosting machine.The Annals of Statistics, 2001
- Long Short-Term MemoryNeural Computation, 1997
- Cross-Validation of Regression ModelsJournal of the American Statistical Association, 1984