Microclimate-Based Frost Prediction Model Resolving the Class Imbalance 


Vol. 47,  No. 10, pp. 1704-1715, Oct.  2022
10.7840/kics.2022.47.10.1704


PDF
  Abstract

Microclimate has been influencing all forms of time-sensitive agriculture. With substantial advances in emerging and enabling technologies, a vast amount of IoT-based environmental data allows preparation for the adverse impacts by providing helpful information to time-sensitive services. Of particular concern among high-risk weather conditions is nonanticipative frosty damage, affecting agricultural yield significantly. This paper proposes a timely frost prediction model based on machine learning using environmental data. Because of minority information on frost, conventional approaches often suffer from the class imbalance problem with rare labeling data. We address these issues through a frost prediction model using class-balanced data by SMOTE method to environmental datasets collected from IoT stations when predictive service executes. Our experimental results demonstrate that the frost prediction using Random Forest is the most suitable algorithm. With the optimization process, the performance of the frost prediction model was improved by about 4% (Based on f1). Moreover, the performance evaluation by SMOTE ratio shows the importance of an appropriate ratio for data augmentation by unique tendency.

  Statistics
Cumulative Counts from November, 2022
Multiple requests among the same browser session are counted as one view. If you mouse over a chart, the values of data points will be shown.


  Related Articles
  Cite this article

[IEEE Style]

A. Moon and H. Kim, "Microclimate-Based Frost Prediction Model Resolving the Class Imbalance," The Journal of Korean Institute of Communications and Information Sciences, vol. 47, no. 10, pp. 1704-1715, 2022. DOI: 10.7840/kics.2022.47.10.1704.

[ACM Style]

Ae-Kyeung Moon and Hyo-Seon Kim. 2022. Microclimate-Based Frost Prediction Model Resolving the Class Imbalance. The Journal of Korean Institute of Communications and Information Sciences, 47, 10, (2022), 1704-1715. DOI: 10.7840/kics.2022.47.10.1704.

[KICS Style]

Ae-Kyeung Moon and Hyo-Seon Kim, "Microclimate-Based Frost Prediction Model Resolving the Class Imbalance," The Journal of Korean Institute of Communications and Information Sciences, vol. 47, no. 10, pp. 1704-1715, 10. 2022. (https://doi.org/10.7840/kics.2022.47.10.1704)

Search
(IN TITLE, AUTHOR, ABSTRACT,KEYWORDS)

Advanced Search ▼

POPULAR KEYWORDS
(TOP 10 KEYWORDS)

Recent Publications
(LAST 3 YEARS)






pISSN : 1226 - 4717
eISSN : 2287 - 3880