All Issue

2024 Vol.33, Issue 4 Preview Page

Original Articles

Comparison of Growth Prediction Models for Pepper Seedling Production Using Monthly Dummy Variables

월별 더미변수를 활용한 고추 공정육묘 생육예측 모델 비교

Yeon Jin Cho1*
,
Seo Woo Jang2
,
Hee Gon Kim2
,
Youn Sup Cho2
조 연진1*
,
장 서우2
,
김 희곤2
,
조 윤섭2
1Researcher, Horticultural Research Institute, Jeollanam-do Agricultural Research & Extension Services, Naju 59213, Korea
2Senior Researcher, Horticultural Research Institute, Jeollanam-do Agricultural Research & Extension Services, Naju 59213, Korea
1전라남도농업기술원 원예연구소 농업연구사
2전라남도농업기술원 원예연구소 농업연구관
*Corresponding Author
31 October 2024. pp. 427-435
PDF XML Citation
Abstract

It’s been said that seedlings cultivation is half of farming. The environment and growth management during the seedling cultivation affect the planting period. Accordingly, a growth prediction model according to environmental changes is required. In this study, we aimed to predict the ‘Height of Plantlet’ of peppers according to environmental changes. To conduct the study, environmental and growth data was collected from september 2021 to 2023 in Jeollanam-do, and processed the data to select five major environmental variables as input variables and the growth variable ‘Height of plantlet’ as an output variable to conduct multiple regression analysis. However, this did not reflect seasonality, so the predictive power was low at 0.72, and the MAPE was 40.5%, showing a large error between the actual and predicted values. To overcome this, we added a ‘Monthly dummy coefficient’ as a seasonal factor and applied multiple regression analysis, random forest, and CatBoost models to compare the performance. Through this, the CatBoost model with the highest explanatory power of 0.96 and the lowest MAPE of 13.47% was selected. This study is expected to contribute to the establishment of a seedling DB and systematic seedling quality management by automatically predicting the ‘chojang’ of peppers using environmental information.

Keywords
dummy coefficient
growth model
growth prediction
red pepper ‘Cheongyang’
seedling

육묘는 농사의 절반이란 말이 있다. 모든 생물이 그렇듯 식물의 초기 발육 단계는 전 생육주기를 결정하는 중요한 요소이다. 우량묘 생산을 위한 환경 및 생육관리가 더 중요해지고, 환경 변화에 따른 생육 예측이 필요해졌다. 따라서 이번 연구에서는 환경 변화에 따른 고추의 주요 생육지표인 ‘초장’을 예측하고자 하였다. 연구 수행을 위해 전라남도농업기술원에서 2021년 9월부터 2023년까지 고추 육묘를 진행하면서 환경 및 생육데이터를 수집하였고, 데이터의 가공을 통해 적산온도와 적산일사량, 일최고온도, 일최저온도 등을 12개 변수를 산출하였다. 생육 데이터간의 상관분석을 통해 시계열적 변화를 이루며 생육기관별 연관성이 높은 ‘초장’을 출력변수로 선택하고, 변수선택법(RFE)를 통해 주요 환경변수 5개를 입력변수로 선정하여 다중회귀분석을 진행하였다. 그러나 이는 계절성을 반영하지 못해 예측 설명력이 0.72로 낮고, MAPE 40.5%로 실측치와 예측치와의 오차가 크게 나타났다. 이를 극복하기 위해 계절적 요인인 ‘월별 더미변수’를 추가하여 다중회귀분석, 랜덤포레스트, 캣부스트 모델을 적용해 성능을 비교하였다. 이를 통해 실제 ‘초장’생육과 예측 생육량을 비교해 설명력(R2)이 0.96으로 가장 높고 MAPE가 13.47%로 가장 낮은 캣부스트 모델을 선정하였다. 4차 산업혁명이 농업에 적용되며 전 분야에서 ‘표준화’와 ‘자동화’가 요구되고 있는데 이번 연구를 통해 환경정보 활용 고추 ‘초장’을 자동으로 예측함으로써 육묘 DB 구축 및 체계적 유묘 품질관리에 기여할 수 있을 것으로 보인다.

키워드
고추 ‘청양’
육묘
생장 모델
더미계수
생육예측
References
1

Garavaglia S., and A. Sharma 1997, A smart guide to dummy variables: Four applications and a macro. Proceedings of the Northeast SAS Users Group Conference. Murray Hill, NJ: Dun & Bradstreet.

2

Guera O. G. M., F. Castrejón-Ayala, N. Robledo, A. Jiménez-Pérez, and L. Sarazar-Marcial 2023, Dummy variable regression and artificial neural networks for maize grain yields prediction in push-pull systems. Research Square. Available via https://www.researchsquare.com Accessed 2 February 2023.

10.21203/rs.3.rs-2521610/v1
3

Gupta I., S. Ayalasomayajula, Y. Shashidhara, A. Kataria, S. Shashidhara, and K. Kataria 2023, Innovations in agricultural forecasting: A multivariate regression study on global crop yield prediction. arXiv preprint arXiv:2312.02254.

10.17148/IJARCCE.2024.13922
4

Hara P., M. Piekutowska, and G. Niedbała 2021, Selection of independent variables for crop yield prediction using artificial neural network models with remote sensing data. Land 10:609.

10.3390/land10060609
5

Jeong J.H., J.P. Resop, N.D. Mueller, D.H. Fleisher, K. Yun, E.E. Butler, D.J. Timlin, K.-M. Shim, J.S. Gerber, V.R. Reddy, and S.-H. Kim 2016, Random forest for global and regional crop yield prediction. PLOS ONE 11:e0156571. doi:10.1371/journal.pone.0156571

10.1371/journal.pone.015657127257967PMC4892571
6

Kim J. S. 2015, Present studies and development plan of seedling industry in Korea. Master's thesis, Konkuk University Graduate School of Agriculture and Life Sciences.

7

Kim S. G., J. H. Lee, H. J. Lee, S. G. Lee, B. H. Moon, S. W. Ahn, and H. S. Lee 2018, Development of prediction growth and yield models by growing degree days in hot pepper. J Bio-Env Con 27:424-430 (in Korean) doi:10.12791/KSBEC.2018.27.4.424

10.12791/KSBEC.2018.27.4.424
8

Korea Rural Economic Institute (KREI). 2011, A study on the current state and development strategies of raising seedlings industry, pp 15-17.

9

Kwack Y., S. W. An, and S. K. Kim 2021, Development of growth model for grafted hot pepper seedlings as affected by air temperature and light intensity. Sustainability 13:5895. doi:10.3390/su13115895

10.3390/su13115895
10

Lobell D.B., and G.P. Asner 2003, Climate and management contributions to recent trends in U.S. agricultural yields. Science 299:1032. doi:10.1126/science.1078475

10.1126/science.107847512586935
11

Matsumura K., C.F. Gaitan, K. Sugimoto, A.J. Cannon, and W. Hsieh 2015, Maize yield forecasting by linear regression and artificial neural networks in Jilin, China. Journal of Agricultural Science 153:399-410. doi:10.1017/S002185 9614000392

10.1017/S0021859614000392
12

Muraina I.O. 2022, Ideal dataset splitting ratios in machine learning algorithms: General concerns for data scientists and data analysts. International Journal of Education and Learning Systems 7:496-504.

13

Rodríguez F., M. Berenguel, J. Guzmán, and A. Ramírez-Arias, 1999, Modeling and control of greenhouse crop growth. Springer, Cham, Switzerland.

14

Rural Development Administration (RDA) 2008, Vegetable Seedling Cultivation Technology, R. Korea, pp 16-19. (in Korean)

15

Rural Development Administration (RDA) 2012, Agricultural Science and Technology Research and Analysis Standards, R. Korea, pp 547-549. (in Korean)

16

Rural Development Administration (RDA) 2017, Pepper Agricultural Technology Guide, R. Korea, 8, pp 60-64. (in Korean)

17

Uribeetxebarria A., A. Castellón, and A. Aizpurua 2023, Optimizing wheat yield prediction integrating data from Sentinel-1 and Sentinel-2 with CatBoost algorithm. Remote Sensing 15:1640. doi:10.3390/rs15061640

10.3390/rs15061640
18

Weng J. 2021, Data Splitting for Model Evaluation. Towards Data Science. Available via https://towardsdatascience.com Accessed 15 October 2021

19

Wildt A.R. 1977, Estimating models of seasonal market response using dummy variables. Journal of Marketing Research 14:34-41. doi:10.1177/002224377701400106

10.1177/002224377701400106
Information
  • Publisher :The Korean Society for Bio-Environment Control
  • Publisher(Ko) :(사)한국생물환경조절학회
  • Journal Title :Journal of Bio-Environment Control
  • Journal Title(Ko) :생물환경조절학회지
  • Volume : 33
  • No :4
  • Pages :427-435
  • Received Date : 2024-10-02
  • Revised Date : 2024-10-23
  • Accepted Date : 2024-10-27
  • DOI :https://doi.org/10.12791/KSBEC.2024.33.4.427
Journal Informaiton Journal of Bio-Environment Control Journal of Bio-Environment Control
Online Submission
  • NRF
  • KOFST
  • crossref crossmark
  • crossref cited-by
  • crossref funder-registry
  • crosscheck
  • orcid
  • open access
  • ccl
Journal Informaiton Journal Informaiton - close