The Journal of Information Systems (한국정보시스템학회지:정보시스템연구)

Korea Association of Information Systems (한국정보시스템학회)
권호 표지
DOI QR코드

DOI QR Code

Exploring process prediction based on deep learning: Focusing on dynamic recurrent neural networks

딥러닝 기반의 프로세스 예측에 관한 연구: 동적 순환신경망을 중심으로

  • Received : 2018.11.16
  • Accepted : 2018.12.15
  • Published : 2018.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

Purpose The purpose of this study is to predict future behaviors of business process. Specifically, this study tried to predict the last activities of process instances. It contributes to overcoming the limitations of existing approaches that they do not accurately reflect the actual behavior of business process and it requires a lot of effort and time every time they are applied to specific processes. Design/methodology/approach This study proposed a novel approach based using deep learning in the form of dynamic recurrent neural networks. To improve the accuracy of our prediction model based on the approach, we tried to adopt the latest techniques including new initialization functions(Xavier and He initializations). The proposed approach has been verified using real-life data of a domestic small and medium-sized business. Findings According to the experiment result, our approach achieves better prediction accuracy than the latest approach based on the static recurrent neural networks. It is also proved that much less effort and time are required to predict the behavior of business processes.

Keywords

JBSTB0_2018_v27n4_115_f0001.png 이미지

<그림 1> 이벤트 로그의 구조

JBSTB0_2018_v27n4_115_f0002.png 이미지

<그림 2> 순환신경망 구조의 예 (2층의 3단계로 펼쳐진 LSTM 셀을 가진 다대다의 RNN 구조)

JBSTB0_2018_v27n4_115_f0003.png 이미지

<그림 3> 정적 순환신경망을 이용한 학습과 예측을 위한 첫 번째 데이터 변환 방법

JBSTB0_2018_v27n4_115_f0004.png 이미지

<그림 4> 정적 순환신경망을 이용한 학습과 예측을 위한 두 번째 데이터 변환 방법

JBSTB0_2018_v27n4_115_f0005.png 이미지

<그림 5> 동적 순환신경망을 이용한 구조

JBSTB0_2018_v27n4_115_f0006.png 이미지

<그림 5> 10개 겹의 예측 정확도 변화 (X축: 에포크, Y축: 정확도)

JBSTB0_2018_v27n4_115_f0007.png 이미지

<그림 6> 10개 겹의 비용 변화 (X축: 에포크, Y축: 비용)

<표 1> 활용 데이터 요약

JBSTB0_2018_v27n4_115_t0001.png 이미지

<표 2> 배치 크기에 따른 예측 정확도

JBSTB0_2018_v27n4_115_t0002.png 이미지

<표 3> 10겹 교차검증 적용에 따른 예측 정확도

JBSTB0_2018_v27n4_115_t0003.png 이미지

References

  1. 강영식, 이보경, (경영자와 실무전문가를 위한) 프로세스 마이닝, 한나래 출판사, 2016.
  2. 김대희, 최승완, 곽수영, "딥러닝 기반의 가짜 얼굴 검출," 한국산업정보학회지, 제 23권, 제5호, 2018, pp. 9-17.
  3. 김진백, 김유일, "인공 신경망의 학습에 있어 가중치 변화방법과 은닉층의 노드수가 예측정확성에 미치는 영향," 정보시스템연구, 제9권, 제1호, 2000, pp. 27-44.
  4. 안성만, "딥러닝의 모형과 응용사례," 지능정보연구, 제22권, 제2호, 2016, pp. 127-142. https://doi.org/10.13088/JIIS.2016.22.2.127
  5. 송현정, 이석준, "딥러닝을 활용한 실시간 주식 거래에서의 매매 빈도 패턴과 예측 시점에 관한 연구: KOSDAQ 시장을 중심으로," 정보시스템연구, 제27권, 제3호, 2018, pp. 123-140.
  6. 최희열, 민윤홍, "딥러닝 소개 및 주요 이슈," 정보처리학회지, 제22권, 제1호, 2015, pp. 7-21.
  7. 트란 광 카이, 송사광, "딥러닝 기반 침수 수위예측: 미국 텍사스 트리니티강 사례연구," 정보과학회논문지, 제44권, 제6호, 2017, pp. 607-612.
  8. 한진영, 조철현, 손인수, "기업의 빅데이터 활용에 관한 실증연구 : A 쇼핑사의 빅데이터기반 통합로그 시스템 사례," 인터넷전자상거래연구, 제15권, 제6호, 2016, pp. 1-19.
  9. Bengio, Y., Simard, P., and Frasconi, P., "Learning long-term dependencies with gradient descent is difficult," Journal of IEEE Transactions on Neural Networks, Vol. 5, No. 2, 1994, pp. 157-166. https://doi.org/10.1109/72.279181
  10. Breuker, D., Matzner, M., Delfmann, P., and Becker, J., "Comprehensible predictive models for business process," MIS Quarterly, Vol. 40, No. 4, 2016, pp. 1009-1034. https://doi.org/10.25300/MISQ/2016/40.4.10
  11. Chung, J., Gulcehre, C., Cho, K., and Bengio, Y., Empirical evaluation of gated recurrent neural networks on sequence modeling, CoRR, abs/1412.3555, 2014.
  12. Evermann, J., Rehse, J.-R., and Fettke, P., "A deep learning approach for predicting process behavior at runtime," PRAISE Workshop at the 14th International Conference on BPM, 2016.
  13. Evermann, J., Rehse, J.-R., and Fettke, P., "Predicting process behaviour using deep learning," Decision Support Systems, Vol. 100, 2017a, pp. 129-140. https://doi.org/10.1016/j.dss.2017.04.003
  14. Evermann, J., Rehse, J.-R., and Fettke, P., "XES TensorFlow-Process prediction using the TensorFlow deep-learning framework," Forum of the Conference on Advanced Information Systems Engineering(CAiSE), Essen, Germany, 2017b.
  15. Glorot, X. and Bengio, Y., "Understanding the difficulty of training deep feedforward neural networks," International Conference on Artificial Intelligence and Statistics, 2010.
  16. Graves, A., "Supervised Sequence Labelling with Recurrent Neural Networks," Berlin Heidelberg, Springer, 2012.
  17. Hinton, G. and Salakhutdinov, R., "Reducing the dimensionality of data with neural networks," Science, Vol. 313, No. 5786, pp. 504-507, 2006. https://doi.org/10.1126/science.1127647
  18. Hochreiter, S. and Schmidhuber, J., "Long short-term memory," Neural Computation, Vol. 9, No. 8, 1997, pp. 1735-1780. https://doi.org/10.1162/neco.1997.9.8.1735
  19. Jung, H. I., Park, I. S., and Ahn, H., "Identifying the key success factors of massively mutiplayer online role playing game design using artificial neural networks," The Journal of Society for e-Business Studies, Vol. 17, No. 1, 2012, pp. 23-38. https://doi.org/10.7838/JSEBS.2012.17.1.023
  20. Kang, Y. S. and Park, S., "Predicting the relationship between corporate financial information and credit rating using deep learning," Korean Journal of Business Administration, Vol. 31, No. 7, 2018, pp. 1253-1275.
  21. Lakshmanan, G., Shamsi, D., Doganata, Y. N., Unuvar, M., and Khalaf, R., "A markov prediction model for data-driven semi-structured business processes," Knowledge and Information Systems, Vol. 42, No. 1, 2015, pp. 97-126. https://doi.org/10.1007/s10115-013-0697-8
  22. LaValle, S., Lesser, E., Shockley, R., Hopkins, M. S., and Kruschwitz, N., "Big data, analytics and the path from insights to value," MIT Sloan Management Review, Vol. 52, 2011, pp. 21-32.
  23. LeCun, Y., Bengio, Y., and Hinton, G., "Deep learning," Nature, Vol. 521, pp. 436-444. 2015. https://doi.org/10.1038/nature14539
  24. Schmidhuber, J., "Deep learning in neural networks: An overview," Neural Networks, Vol. 61, 2015, pp. 85-117. https://doi.org/10.1016/j.neunet.2014.09.003
  25. Sutskever, I., Martens, J., and Hinton, G. E., "Generating text with recurrent neural networks," ICML, Omnipress, 2011, pp. 1017-1024.
  26. Tax, N., Verenich, I., Rosa, M. L., and Dumas, M., "Predictive business process monitoring with LSTM neural networks," CoRR, abs/1612.02130, 2016.
  27. Unuvar, M., Lakshmanan, G. T., and Doganata, Y. N., "Leveraging path information to generate predictions for parallel business processes," Knowledge and Information Systems, Vol. 47, No. 2, 2016, pp. 433-461. https://doi.org/10.1007/s10115-015-0842-7
  28. van der Aalst, W. M. P., "Process Mining: Data Science in Action," Springer, 2016.
  29. van der Aalst, W. M. P., Schonenberg, M. H., and Song, M., "Time prediction based on process mining," Information Systems, Vol. 36, No. 2, 2011, pp. 450-475. https://doi.org/10.1016/j.is.2010.09.001