화학공학소재연구정보센터
홈 로그인 로그아웃 연락처 사이트맵
Korean Chemical Engineering Research, Vol.59, No.4, 532-541, November, 2021
DOI10.9713/kcer.2021.59.4.532  Export Citation
시계열 교차검증을 적용한 2,3-BDO 분리공정 온도예측 모델의 초매개변수 최적화
Application of Time-series Cross Validation in Hyperparameter Tuning of a Predictive Model for 2,3-BDO Distillation Process
안나현1, 2, 최영렬1, 2, 조형태1, 김정환1, †
  • 1한국생산기술연구원 친환경재료공정연구그룹, 44413 울산광역시 중구 종가로 55
  • 2연세대학교 화공생명공학과, 03722 서울특별시 서대문구 연세로 50
Nahyeon An1, 2, Yeongryeol Choi1, 2, Hyungtae Cho1, and Junghwan Kim1, †
  • 1Green Materials and Processes R&D Group, Korea Institute of Industrial Technology, 55, Jongga-ro, Ulsan, 44413, Korea
  • 2Department of Chemical and Biomolecular Engineering, Yonsei University, 50, Yonsei-ro, Seoul, 03722, Korea
E-mail:
초록
최근 인공지능에 대한 관심이 높아짐에 따라 화학공정분야에서도 인공지능을 활용한 연구가 많아지고 있다. 그러나 인공지능 기반 모델이 충분히 일반화되지 않아 학습에 이용되지 않은 새로운 데이터에 대한 예측률이 떨어지는 과적합 현상이 빈번하게 일어나고 있으며, 교차검증은 과적합을 해결하는 방법 중 하나이다. 본 연구에서는 2,3-BDO 분리공정 온도 예측 모델의 초매개변수 중에서 배치 개수와 반복횟수를 조정하기 위해 시계열 교차검증을 적용하고 일반적으로 사용되는 K 겹 교차검증과 비교하였다. 결과적으로 K 겹 교차검증을 사용했을 때 보다 시계열 교차검증 방식을 사용했을 때 MAPE는 0.61% 증가한 반면 RMSE는 9.06% 감소하였고 학습 시간은 198.29초 적게 소요되었다.
Recently, research on the application of artificial intelligence in the chemical process has been increasing rapidly. However, overfitting is a significant problem that prevents the model from being generalized well to predict unseen data on test data, as well as observed training data. Cross validation is one of the ways to solve the overfitting problem. In this study, the time-series cross validation method was applied to optimize the number of batch and epoch in the hyperparameters of the prediction model for the 2,3-BDO distillation process, and it compared with K-fold cross validation generally used. As a result, the RMSE of the model with time-series cross validation was lower by 9.06%, and the MAPE was higher by 0.61% than the model with K-fold cross validation. Also, the calculation time was 198.29 sec less than the K-fold cross validation method.
Keywords:Cross validation;Distillation process;Predictive model;Hyperparameter tuning;Time-series cross validation
  1. Oh KC, Kwon HK, Roh JW, Choi YR, Park HD, Cho HT, Kim JH, Korean Chem. Eng. Res., 58(4), 565 (2020)
  2. Hoon S, Ah Y, Hyeong J, J. Korean Soc. Qual. Manag., 48(3), 499 (2020)
  3. Zhai Y, Yao P, Zhou X, IEEE, ITAIC 2020 - IEEE 9th Joint International Information Technology and Artificial Intelligence Conferencepp. 1397-1400.
  4. Lee YC, Choi YR, Cho HT, Kim JH, Korean Chem. Eng. Res., 59(2), 191 (2021)
  5. Lu ZJ, Xiang Q, Wu YM, Gu J, IEEE Int. Conf. Ind. Informatics, INDIN 2015, 98 (2015).
  6. Wu H, Zhao JS, Comput. Chem. Eng., 115, 185 (2018)
  7. Eslamloueyan R, Appl. Soft Comput. J., 11(1), 1407 (2011)
  8. Wei YQ, Weng ZX, Can. J. Chem. Eng., 98(6), 1293 (2020)
  9. Jing C, Hou J, Neurocomputing, 167, 636 (2015)
  10. Wang T, Gao H, Qiu J, IEEE Trans. Neural Networks Learn. Syst., 27(2), 416 (2016)
  11. Mazinan AH, Int. J. Adv. Manuf. Technol., 66(9-12), 1379 (2013)
  12. Mahdi M, Mehdi B, Springer Netherlands (2021).
  13. Arlot S, Celisse A, Stat. Surv., 4, 40 (2010)
  14. Srivastava N, Hinton G, Krizhevsky A, Sutskever I, Salakhutdinov R, J. Mach. Learn. Res.,(2014).
  15. Ying X, J. Phys. Conf. Ser., 1168(2), 2019
  16. Bergmeir C, Benitez JM, Inf. Sci. (Ny)., 191, 192 (2012)
  17. Chen X, Chen X, She J, Wu M, Neurocomputing, (2017).
  18. Zhao J, Wang W, Sheng C, Data-driven prediction for industrial processes and their applications, (2018).
  19. Benesty J, Chen J, Huang Y, Cohen I, Pearson Correlation Coefficient, (2009).
  20. Andreas CM, Sarah G, thirdIntroduction to machine learning with python, O’Reilly(2020).
  21. Hochreiter S, Schmidhuber JU, “Long Shortterm Memory,” Neural Comput., (1997).
  22. Brownlee J, Mach. Learn. Mastery, (2018).
  23. Kingma DP, Ba JL, 3rd Int. Conf. Learn. Represent. ICLR 2015 - Conf. Track Proc.