Scheduled refitting can replace an effective customer segmentation even when a newly fitted model offers little improvement. We examine selective replacement for batchwise customer segmentation under concept drift, comparing BIRCH and MiniBatchKMeans under static, periodic, quality-only, and reward-based policies on Instacart, Ta-Feng, and Olist. The reward policy has the highest reported mean deployed silhouette in all six dataset-algorithm combinations; its advantage over quality-only is 0.0045 - 0.0170 silhouette points. Five supplied paired seed effects per combination are positive for reward versus each baseline, with seed-bootstrap intervals excluding zero. These same-snapshot comparisons do not establish temporally disjoint gains or isolate the reward’s stability term. The raw run record is not yet archived; label-aligned stability and denominator-aware event analyses are not reported.
References
[1]
Gama, J., Žliobaitė, I., Bifet, A., Pechenizkiy, M. and Bouchachia, A. (2014) A Survey on Concept Drift Adaptation. <i>ACM Computing Surveys</i>, 46, 1-37. <br>https://doi.org/10.1145/2523813
[2]
Fader, P.S., Hardie, B.G.S. and Lee, K.L. (2005) RFM and CLV: Using Iso-Value Curves for Customer Base Analysis. <i>Journal of Marketing Research</i>, 42, 415-430. <br>https://doi.org/10.1509/jmkr.2005.42.4.415
[3]
Jain, A.K. (2010) Data Clustering: 50 Years beyond K-Means. <i>Pattern Recognition Letters</i>, 31, 651-666. <br>https://doi.org/10.1016/j.patrec.2009.09.011
[4]
Zhang, T., Ramakrishnan, R. and Livny, M. (1996) BIRCH: An Efficient Data Clustering Method for Very Large Databases. <i>Proceedings of the</i> 1996 <i>ACM SIGMOD International Conference on Management of Data—SIGMOD</i>’96, Montreal, 4-6 June 1996, 103-114. <br>https://doi.org/10.1145/233269.233324
[5]
Sculley, D. (2010) Web-Scale K-Means Clustering. <i>Proceedings of the</i> 19<i>th International Conference on World Wide Web</i>, Raleigh, 26-30 April 2010, 1177-1178. <br>https://doi.org/10.1145/1772690.1772862
[6]
Dawid, A.P. (1984) Present Position and Potential Developments: Some Personal Views: Statistical Theory: The Prequential Approach. <i>Journal of the Royal Statistical Society. Series </i><i>A</i> (<i>General</i>), 147, 278. <br>https://doi.org/10.2307/2981683
[7]
Rousseeuw, P.J. (1987) Silhouettes: A Graphical Aid to the Interpretation and Validation of Cluster Analysis. <i>Journal of Computational and Applied Mathematics</i>, 20, 53-65. <br>https://doi.org/10.1016/0377-0427(87)90125-7
[8]
Hubert, L. and Arabie, P. (1985) Comparing Partitions. <i>Journal of Classification</i>, 2, 193-218. <br>https://doi.org/10.1007/bf01908075
[9]
Aggarwal, C.C., Yu, P.S., Han, J. and Wang, J. (2003) A Framework for Clustering Evolving Data Streams. In: Freytag, J.C., et al., Eds., <i>Proceedings </i>2003<i> VLDB Conference</i>, Elsevier, 81-92. <br>https://doi.org/10.1016/b978-012722442-8/50016-1
[10]
Cao, F., Estert, M., Qian, W. and Zhou, A. (2006) Density-Based Clustering over an Evolving Data Stream with Noise. <i>Proceedings of the</i> 2006 <i>SIAM International Conference on Data Mining</i>, Bethesda, 20-22 April 2006, 328-339. <br>https://doi.org/10.1137/1.9781611972764.29
[11]
Chen, Y. and Tu, L. (2007) Density-Based Clustering for Real-Time Stream Data. <i>Proceedings of the</i> 13<i>th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining</i>, San Jose, 12-15 August 2007, 133-142. <br>https://doi.org/10.1145/1281192.1281210
[12]
Kranen, P., Assent, I., Baldauf, C. and Seidl, T. (2010) The ClusTree: Indexing Micro-Clusters for Anytime Stream Mining. <i>Knowledge and Information Systems</i>, 29, 249-272. <br>https://doi.org/10.1007/s10115-010-0342-8
[13]
Ackermann, M.R., Märtens, M., Raupach, C., Swierkot, K., Lammersen, C. and Sohler, C. (2012) StreamKM++: A Clustering Algorithm for Data Streams. <i>ACM Journal of Experi</i><i>mental Algorithmics</i>, 17, 2.1-2.30. <br>https://doi.org/10.1145/2133803.2184450
[14]
Wang, X., Wang, Z., Wu, Z., Zhang, S., Shi, X. and Lu, L. (2023) Data Stream Clustering: An In-Depth Empirical Study. <i>Proceedings of the ACM on </i><i>Management of Data</i>, 1, 1-26. <br>https://doi.org/10.1145/3589307
[15]
Wang, Z., Wang, X. and Zhang, S. (2024) MOStream: A Modular and Self-Optimizing Data Stream Clustering Algorithm. 2024 <i>IEEE International Conference on Data Mining</i> (<i>ICDM</i>), Abu Dhabi, 9-12 December 2024,500-509. <br>https://doi.org/10.1109/icdm59182.2024.00057
[16]
Bifet, A. and Gavaldà, R. (2007) Learning from Time-Changing Data with Adaptive Windowing. <i>Proceedings of the</i> 2007 <i>SIAM International Conference on Data Mining</i>, Minneapolis, 26-28 April 2007, 443-448. <br>https://doi.org/10.1137/1.9781611972771.42
[17]
Gretton, A., Borgwardt, K.M., Rasch, M.J., Schölkopf, B. and Smola, A.J. (2012) A Kernel Two-Sample Test. <i>Journal of Machine Learning Research</i>, 13, 723-773.
[18]
Hennig, C. (2007) Cluster-Wise Assessment of Cluster Stability. <i>Computational Statistics & Data Analysis</i>, 52, 258-271. <br>https://doi.org/10.1016/j.csda.2006.11.025
[19]
Bifet, A., Holmes, G., Kirkby, R. and Pfahringer, B. (2010) MOA: Massive Online Analysis. <i>Journal of Machine Learning Research</i>, 11, 1601-1604.
[20]
(2017) InstaCart Online Grocery Basket Analysis Dataset. Kaggle, Dataset 1889830 (Mirror of the 2017 Instacart Release).<br>https://www.kaggle.com/datasets/yasserh/instacart-online-grocery-basket-analysis-dataset/data
[21]
(2026) Ta Feng Grocery Dataset. Kaggle, Dataset 22987. <br>https://www.kaggle.com/datasets/chiranjivdas09/ta-feng-grocery-dataset/data
[22]
Olist (2026) Brazilian E-Commerce Public Dataset by Olist. Kaggle, Dataset 55151. <br>https://www.kaggle.com/datasets/olistbr/brazilian-ecommerce
[23]
Pedregosa, F., et al. (2011) Scikit-Learn: Machine Learning in Python. <i>The Journal of Machine Learning Research</i>, 12, 2825-2830.
