Can machine learning classification methods improve the prediction of leaf wetness in North-Western Europe compared to established empirical methods?

作者：

Highlights：

• Machine learning (ML) outperformed empirical models for prediction of leaf wetness.

• ML models were significantly more accurate than empirical models for most subsets.

• On a regional basis ML were up to 8.78% more accurate than empirical methods.

• The night-time subset of data is the least accurately predicted of all data subsets.

摘要

•Machine learning (ML) outperformed empirical models for prediction of leaf wetness.•ML models were significantly more accurate than empirical models for most subsets.•On a regional basis ML were up to 8.78% more accurate than empirical methods.•The night-time subset of data is the least accurately predicted of all data subsets.

论文关键词：Leaf Wetness,Machine Learning,Classification,Meteorology,Agriculture,Crop Disease

论文评审过程：Received 31 March 2020, Revised 4 March 2021, Accepted 17 May 2021, Available online 23 May 2021, Version of Record 27 May 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.115255