Korean Chemical Engineering Research

The Korean Institute of Chemical Engineers (한국화학공학회)
권호 표지
DOI QR코드

DOI QR Code

Application of Time-series Cross Validation in Hyperparameter Tuning of a Predictive Model for 2,3-BDO Distillation Process

시계열 교차검증을 적용한 2,3-BDO 분리공정 온도예측 모델의 초매개변수 최적화

  • An, Nahyeon (Green Materials and Processes R&D Group, Korea Institute of Industrial Technology) ;
  • Choi, Yeongryeol (Green Materials and Processes R&D Group, Korea Institute of Industrial Technology) ;
  • Cho, Hyungtae (Green Materials and Processes R&D Group, Korea Institute of Industrial Technology) ;
  • Kim, Junghwan (Green Materials and Processes R&D Group, Korea Institute of Industrial Technology)
  • 안나현 (한국생산기술연구원 친환경재료공정연구그룹) ;
  • 최영렬 (한국생산기술연구원 친환경재료공정연구그룹) ;
  • 조형태 (한국생산기술연구원 친환경재료공정연구그룹) ;
  • 김정환 (한국생산기술연구원 친환경재료공정연구그룹)
  • Received : 2021.05.31
  • Accepted : 2021.08.05
  • Published : 2021.11.01
Download PDF
⟨ Previous Next ⟩

Abstract

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.

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 분리 공정 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.

Keywords

Acknowledgement

본 논문은 한국생산기술연구원 "기업체 에너지공정 최적화 지원 사업(EM-21-0022)" 및 "화학산업 고도화를 위한 스마트 제조공정 AI 플랫폼 기술 개발(JH-21-0005)"의 지원으로 수행한 연구입니다.

References

  1. Oh, K. C., Kwon, H., Roh, J., Choi, Y., Park, H., Cho, H. and Kim, J., "Development of Machine Learning-Based Platform for Distillation Column," Korean Chem. Eng. Res., 58(4), 565-572 (2020).
  2. Hoon, S., Ah, Y. and Hyeong, J., "A Machine Learning Model for Predicting Silica Concentrations through Time Series Analysis of Mining Data," J. Korean Soc. Qual. Manag., 48(3), 499-508(2020). https://doi.org/10.7469/JKSQM.2020.48.3.499
  3. Zhai, N., Yao, P. and Zhou, X., "Multivariate Time Series Forecast in Industrial Process Based on XGBoost and GRU," in, IEEE, ITAIC 2020 - IEEE 9th Joint International Information Technology and Artificial Intelligence Conferencepp. 1397-1400.
  4. Lee, Y., Choi, Y., Cho, H. and Kim, J., "Prediction of Distillation Column Temperature Using Machine Learning and Data Preprocessing," Korean Chem. Eng. Res., 59(2), 191-199(2021).
  5. Lu, Z. J., Xiang, Q., Wu, Y. M. and Gu, J., "Application of Support Vector Machine and Genetic Algorithm Optimization for Quality Prediction Within Complex Industrial Process," Proceeding - 2015 IEEE Int. Conf. Ind. Informatics, INDIN 2015, 98-103(2015).
  6. Wu, H. and Zhao, J., "Deep Convolutional Neural Network Model Based Chemical Process Fault Diagnosis," Comput. Chem. Eng., 115, 185-197(2018). https://doi.org/10.1016/j.compchemeng.2018.04.009
  7. Eslamloueyan, R., "Designing a Hierarchical Neural Network Based on Fuzzy Clustering for Fault Diagnosis of the Tennessee-Eastman Process," Appl. Soft Comput. J., 11(1), 1407-1415(2011). https://doi.org/10.1016/j.asoc.2010.04.012
  8. Wei, Y. and Weng, Z., "Research on TE Process Fault Diagnosis Method Based on DBN and Dropout," Can. J. Chem. Eng., 98(6), 1293-1306(2020). https://doi.org/10.1002/cjce.23750
  9. Jing, C. and Hou, J., "SVM and PCA Based Fault Classification Approaches for Complicated Industrial Process," Neurocomputing, 167, 636-642(2015). https://doi.org/10.1016/j.neucom.2015.03.082
  10. Wang, T., Gao, H. and Qiu, J., "A Combined Adaptive Neural Network and Nonlinear Model Predictive Control for Multirate Networked Industrial Process Control," IEEE Trans. Neural Networks Learn. Syst., 27(2), 416-425(2016). https://doi.org/10.1109/TNNLS.2015.2411671
  11. Mazinan, A. H., "A New Algorithm to AI-based Predictive Control Scheme for a Distillation Column System," Int. J. Adv. Manuf. Technol., 66(9-12), 1379-1388(2013). https://doi.org/10.1007/s00170-012-4415-6
  12. Mahdi, M. and Mehdi, B., A systematic review on overfitting control in shallow and deep neural networks, Springer Netherlands(2021).
  13. Arlot, S. and Celisse, A., "A Survey of Cross-validation Procedures for Model Selection," Stat. Surv., 4, 40-79(2010). https://doi.org/10.1214/09-SS054
  14. Srivastava, N., Hinton, G., Krizhevsky, A., Sutskever, I. and Salakhutdinov, R., "Dropout: A Simple Way to Prevent Neural Networks from Overfitting," J. Mach. Learn. Res.,(2014).
  15. Ying, X., "An Overview of Overfitting and its Solutions," J. Phys. Conf. Ser., 1168(2), (2019).
  16. Bergmeir, C. and Benitez, J. M., "On the Use of Cross-validation for Time Series Predictor Evaluation," Inf. Sci. (Ny)., 191, 192-213(2012). https://doi.org/10.1016/j.ins.2011.12.028
  17. Chen, X., Chen, X., She, J. and Wu, M., "A Hybrid Time Series Prediction Model Based on Recurrent Neural Network and Double Joint Linear-nonlinear Extreme Learning Network for Prediction of Carbon Efficiency in Iron Ore Sintering Process," Neurocomputing, (2017).
  18. Zhao, J., Wang, W. and Sheng, C., Data-driven prediction for industrial processes and their applications, (2018).
  19. Benesty, J., Chen, J., Huang, Y. and Cohen, I., Pearson Correlation Coefficient, (2009).
  20. Andreas C. M. and Sarah, G., thirdIntroduction to machine learning with python, O'Reilly(2020).
  21. Hochreiter, S. and Urgen Schmidhuber, J., "Long Shortterm Memory," Neural Comput., (1997).
  22. Brownlee, J., "What is the Difference Between a Batch and an Epoch in a Neural Network?," Mach. Learn. Mastery, (2018).
  23. Kingma, D. P. and Ba, J. L., "Adam: A Method for Stochastic Optimization," 3rd Int. Conf. Learn. Represent. ICLR 2015 - Conf. Track Proc.