A NEW APPROACH IN INTELLIGENT MAINTENANCE: SPARE PART PREDICTION ON REAL TIME BASIS

Öz

After sales customer services attempt to satisfy target customers by qualified spare part services and affordable prices. While undesirable repetitive customer visits result in significant increase at sales operational costs, this misleading policy may deteriorate organizational goodwill at long run. In the context of intelligent maintenance, spare part prediction solution proactively proposes the most probable spare part to the responsible technical service by a hybrid classification model that combines C4.5, Apriori algorithms and weighted k-Nearest Neighbor (kNN) adaptations. The experimental results demonstrate that proposed approach approximately doubles the human-level performance at spare part prediction scenario. According to best runtime configuration analysis, a real time spare part prediction model has been deployed for television product group at the client’s SAP system.

Anahtar Kelimeler

• Classification
• Intelligent Maintenance
• SAP
• Spare Part Prediction on Real Time Basis
• Weighted k-Nearest Neighbor

AKILLI BAKIMDA YENİ BİR YAKLAŞIM: GERÇEK ZAMANLI YEDEK PARÇA TAHMİNLEME

Öz

Satış sonrası müşteri hizmetleri, kaliteli yedek parça hizmetleri ve uygun fiyatlar ile hedef müşteri kitlesini memnun etmeye çalışmaktadır. Sıkça tekrarlayan müşteri ziyaretleri operasyonel satış maliyetlerinde önemli artışlara neden olurken, bu yanıltıcı politika uzun vadede kurumsal itibarı da zedeleyebilir. Akıllı bakım bağlamında gerçek zamanlı yedek parça tahminleme çözümü, C4.5, Apriori algoritmaları ve ağırlıklandırılmış k-en yakın komşuluk (kNN) uyarlamalarını birleştiren hibrit bir sınıflandırma modeli ile sorumlu teknik servise ilgili arıza için en olası yedek parçayı proaktif olarak önerir. Deneysel sonuçlara göre, önerilen yaklaşımın insan düzeyindeki yedek parça tahminleme performansını yaklaşık iki katına çıkardığı görülmektedir. En iyi model konfigürasyonuna göre televizyon ürün grubu için gerçek zamanlı yedek parça tahminleme çözümü SAP sisteminde canlı kullanıma alınmıştır.

Anahtar Kelimeler

• Sınıflandırma
• Akıllı Bakım
• SAP
• Gerçek Zamanlı Yedek Parça Tahmini
• Ağırlıklandırılmış k-En Yakın Komşuluk

Kaynakça

1. Aguilera, J., Gonzalez, L. C., Montes-y-Gomez, M., ve Rosso, P. (2018). A new weighted k-nearest neighbor algorithm based on newton's gravitational force. In Iberoamerican Congress on Pattern Recognition, 305-313. doi:10.1007/978-3-030-13469-3_36.
2. Boriah, S., Chandola, V., ve Kumar, V. (2008). Similarity measures for categorical data: A comparative evaluation, In Proceedings of the 2008 SIAM international conference on data mining, 243-254. doi:10.1137/1.9781611972788.22.
3. Chapman, P., Clinton, J., Kerber, R., Khabaza, T., Reinartz, T., Shearer, C., Wirth, R. (1999). CRISP-DM 1.0 Step-by-Step Data Mining Guide.
4. Domingos, P. (2015). The master algorithm: How the quest for the ultimate learning machine will remake our world. Basic Books.
5. Dudani, S.A. (1978). The distance-weighted k-nearest-neighbor rule. IEEE trans. on systems, man and cybernetics, 8(4), 311-313.
6. Kantardzic, M. (2011). Data mining: concepts, models, methods, and algorithms. John Wiley & So Liu, C., Cao, L. ve Yu, P. S. (2014). A hybrid coupled k-nearest neighbor algorithm on imbalance data. International Joint Conference on Neural Networks (IJCNN), 2011-2018. doi:10.1109/IJCNN.2014.6889798.
7. Liu, W. ve Chawla, S. (2011). Class confidence weighted knn algorithms for imbalanced data sets, In Pacific-Asia conference on knowledge discovery and data mining, 345-356. doi:10.1007/978-3-642-20847-8_29.
8. Mateos-Garcia, D., Garcia-Gutierrez, J., ve Riquelme-Santos, J.C. (2016). An evolutionary voting for k-nearest neighbors. Expert Systems with Applications, 43, 9-14. doi:10.1016/j.eswa.2015.08.017.

9. Pang-Ning, T., Steinbach, M., ve Kumar, V. (2006). Introduction to data mining, Pearson Education India.
10. Parvinnia, E., Sabeti, M., Zolghadri Jahromi, M. ve Boostani, R. (2014). Classification of EEG Signals using adaptive weighted distance nearest neighbor algorithm. Journal of King Saud University-Computer and Information Sciences, 26(1), 1-6. doi:10.1016/j.jksuci.2013.01.001.
11. Song, Y., Huang, J., Zhou, D., Zha, H. ve Giles, C. L. (2007). IkNN: Informative k-nearest neighbor pattern classification, European Conference on Principles of Data Mining and Knowledge Discovery, 248-264. doi:10.1007/978-3-540-74976-9_25.
12. Tan, S. (2005). Neighbor-weighted k-nearest neighbor for unbalanced text corpus. Expert Systems with Applications, 28(4), 667-671. doi:10.1016/j.eswa.2004.12.023.
13. Wang, C., Cao, L., Wang, M., Li, J., Wei, W., ve Ou, Y. (2011). Coupled nominal similarity in unsupervised learning, Proceedings of the 20th ACM international conference on Information and knowledge management, 973-978. doi:10.1145/2063576.2063715.
14. Yuxuan, L. & Zhang, X. (2011). Improving k nearest neighbor with exemplar generalization for imbalanced classification, Pacific-Asia Conference on Knowledge Discovery and Data Mining, 321-332. doi:10.1007/978-3-642-20847-8_27.
15. Zhang, S., Cheng, D., Deng, Z., Zong, M., ve Deng, X. (2018). A novel kNN algorithm with data-driven k parameter computation. Pattern Recognition Letters, 109, 44-54. doi:10.1016/j.patrec.2017.09.036.
16. Zhu, Q., Feng, J., ve Huang, J. (2016). Natural neighbor: A selfadaptive neighborhood method without parameter K. Pattern Recognition Letters, 80, 30-36. doi:10.1016/j.patrec.2016.05.007.