Treffer: An Improved Approach to Prediction of Maize Disease Occurrence Based on Weather Monitoring and Machine Learning: Case of the Forest-Steppe and Northern Steppe of Ukraine.

Agriculture is becoming an increasingly computerised and knowledge-intensive industry in the context of the current global digitalisation and intellectualisation of production processes. This plays a crucial role in ensuring food security at the national and global levels. The subject of the study is software components and computer models for increasing the efficiency of the transformation of weather data based on machine learning. Three types of machine learning algorithms were investigated: Linear regression, Random Forest and Feedforward neural network. The best results were obtained using Random Forest. The main scientific and applied effect of the study in this article is a substantiated approach to improving software and hardware solutions of information technologies for monitoring the soil and climatic conditions of agricultural open-field crop production. This involved the development of software components and computer models for predicting the probability of maize diseases. This effect has been achieved through the implementation of comprehensive studies that include: preliminary approximation of input datasets; identification of models for predictive analytics of the probability of occurrence of specific types of maize diseases in certain agroclimatic zones, taking into account the cumulative impact of climatic parameters and the probability of disease occurrence; formalised accounting of expert experience in the field of crop stress tolerance; the development of software components in the Python programming language that can be integrated into the low-level data processing link. [ABSTRACT FROM AUTHOR]

Copyright of Baltic Journal of Modern Computing is the property of University of Latvia and its content may not be copied or emailed to multiple sites without the copyright holder's express written permission. Additionally, content may not be used with any artificial intelligence tools or machine learning technologies. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)