Adaptive Segmentation of Information Sequences for Machine Learning Modular Regression Models
Downloads
The research objective is to construct an adaptive model for modular machine learning structures that improves the processing quality of information sequences. The novelty of the proposed methodology is that it can identify segments of an information sequence obtained using various methods and assign models with the best quality indicator values to subsequences. Classical methods allow tuning of the model to the entire data sample. The improvement consists of the proposed solutions that consider the inverse problem of forming segments of data sequences, such that their properties correspond to the processing model. The proposed methodology was tested on various models and datasets. Segmentation and assignment of regression models with the best characteristics to individual segments allow the reduction of the mean square error (MSE) and mean absolute error (MAE) to 8%. The findings show an opportunity to increase of 5-8% for weak LR, SVM, and GR models, while strong DT, CNN, ANN, ANFIS, and XGBoost models improve by 1-4% in scenarios with limited data. Segmentation enables more efficient training and reduces sensitivity to noise and outliers. The proposed solution allows the selection of segmentation strategies and model combinations considering local data properties. Its application enables the implementation of existing machine learning architectures to improve the quality of training and analysis of information sequences and increase adaptability, scalability, and interpretability.
Downloads
[1] He, Y., Zhang, X., Kong, X., Yao, L., & Song, Z. (2025). Causality-driven sequence segmentation assisted soft sensing for multiphase industrial processes. Neurocomputing, 631, 129612. doi:10.1016/j.neucom.2025.129612.
[2] Schober, P., & Vetter, T. R. (2021). Segmented Regression in an Interrupted Time Series Study Design. Anesthesia and Analgesia, 132(3), 696–697. doi:10.1213/ANE.0000000000005269.
[3] Pan, Y., Zhang, R., Guo, J., Peng, S., Wu, F., Yuan, K., Gao, Y., Lan, S., Chen, R., Li, L., Hu, X., Du, Z., Zhang, Z., Zhang, X., Li, W., Guo, Q., & Chen, Y. (2025). Morphology generalizable reinforcement learning via multi-level graph features. Neurocomputing, 631, 129644. doi:10.1016/j.neucom.2025.129644.
[4] Britzger, D. (2022). The Linear Template Fit. The European Physical Journal C, 82(8). doi:10.1140/epjc/s10052-022-10581-w.
[5] Jarantow, S. W., Pisors, E. D., & Chiu, M. L. (2023). Introduction to the Use of Linear and Nonlinear Regression Analysis in Quantitative Biological Assays. Current Protocols, 3(6), 801. doi:10.1002/cpz1.801.
[6] Goepp, V., Bouaziz, O., & Nuel, G. (2025). Spline regression with automatic knot selection. Computational Statistics & Data Analysis, 202, 108043. doi:10.1016/j.csda.2024.108043.
[7] Taye, M. M. (2023). Theoretical Understanding of Convolutional Neural Network: Concepts, Architectures, Applications, Future Directions. Computation, 11(3), 52. doi:10.3390/computation11030052.
[8] Motamedi, B., & Villányi, B. (2025). A predictive analytics approach with Bayesian-optimized gentle boosting ensemble models for diabetes diagnosis. Computer Methods and Programs in Biomedicine Update, 7, 100184. doi:10.1016/j.cmpbup.2025.100184.
[9] Beiu, V., & Zawadzki, A. (2005). On Kolmogorov’s superpositions: novel gates and circuits for nanoelectronics? Proceedings. 2005 IEEE International Joint Conference on Neural Networks, 2005., 2, 651–656. doi:10.1109/ijcnn.2005.1555908.
[10] Girosi, F., & Poggio, T. (1989). Representation Properties of Networks: Kolmogorov’s Theorem Is Irrelevant. Neural Computation, 1(4), 465–469. doi:10.1162/neco.1989.1.4.465.
[11] Marques, H. O., Swersky, L., Sander, J., Campello, R. J. G. B., & Zimek, A. (2023). On the evaluation of outlier detection and one-class classification: a comparative study of algorithms, model selection, and ensembles. Data Mining and Knowledge Discovery, 37(4), 1473–1517. doi:10.1007/s10618-023-00931-x.
[12] Pillonetto, G., Aravkin, A., Gedon, D., Ljung, L., Ribeiro, A. H., & Schön, T. B. (2025). Deep networks for system identification: A survey. Automatica, 171, 111907. doi:10.1016/j.automatica.2024.111907.
[13] Rinaldo, A., Wang, D., Wen, Q., Willett, R., & Yu, Y. (2021). Localizing Changes in High-Dimensional Regression Models. Proceedings of Machine Learning Research, 130, 2089–2097. doi:10.48550/arXiv.2010.10410.
[14] Aue, A., Rice, G., & Sönmez, O. (2018). Detecting and dating structural breaks in functional data without dimension reduction. Journal of the Royal Statistical Society. Series B: Statistical Methodology, 80(3), 509–529. doi:10.1111/rssb.12257.
[15] Jin, Y., & Yang, Y. (2025). A survey on knowledge graph-based click-through rate prediction. Expert Systems with Applications, 281, 127501. doi:10.1016/j.eswa.2025.127501.
[16] Belitser, E., & Ghosal, S. (2025). Bayesian uncertainty quantification and structure detection for multiple change points models. Bernoulli, 31(2), 1181–1205. doi:10.3150/24-BEJ1766.
[17] Haynes, K., Fearnhead, P., & Eckley, I. A. (2017). A computationally efficient nonparametric approach for changepoint detection. Statistics and Computing, 27(5), 1293–1305. doi:10.1007/s11222-016-9687-5.
[18] Yoo, N. Y., Lee, H., & Cha, J. H. (2025). Development of a new general class of bivariate distributions based on reversed hazard rate order. Computational Statistics & Data Analysis, 204, 108106. doi:10.1016/j.csda.2024.108106.
[19] Wei, Y., Du, M., Li, T., Zheng, X., & Ji, C. (2024). Feature-fused residual network for time series classification. Journal of King Saud University - Computer and Information Sciences, 36(10), 102227. doi:10.1016/j.jksuci.2024.102227.
[20] Golzari Oskouei, A., Samadi, N., Khezri, S., Najafi Moghaddam, A., Babaei, H., Hamini, K., Fath Nojavan, S., Bouyer, A., & Arasteh, B. (2025). Feature-weighted fuzzy clustering methods: An experimental review. Neurocomputing, 619, 129176. doi:10.1016/j.neucom.2024.129176.
[21] Bardwell, L., & Fearnhead, P. (2017). Bayesian detection of abnormal segments in multiple time series. Bayesian Analysis, 12(1), 193–218. doi:10.1214/16-BA998.
[22] Wang, Y., Rosli, M. M., Musa, N., & Wang, L. (2024). Improving clustering-based and adaptive position-aware interpolation oversampling for imbalanced data classification. Journal of King Saud University - Computer and Information Sciences, 36(10), 102253. doi:10.1016/j.jksuci.2024.102253.
[23] Huang, J., Chen, P., Lu, L., Deng, Y., & Zou, Q. (2023). WCDForest: a weighted cascade deep forest model toward the classification tasks. Applied Intelligence, 53(23), 29169–29182. doi:10.1007/s10489-023-04794-z.
[24] Lu, K. P., & Chang, S. T. (2019). Fuzzy maximum likelihood change-point algorithms for identifying the time of shifts in process data. Neural Computing and Applications, 31(7), 2431–2446. doi:10.1007/s00521-017-3200-8.
[25] Tallman, E., & West, M. (2024). Bayesian predictive decision synthesis. Journal of the Royal Statistical Society. Series B: Statistical Methodology, 86(2), 340–363. doi:10.1093/jrsssb/qkad109.
[26] Mehta, S., Kumar, L., Misra, S., Patnaik, K. S., & Singh, V. (2025). Nested deep learning with learned network embeddings for software defect prediction. Applied Soft Computing, 174, 113057. doi:10.1016/j.asoc.2025.113057.
[27] Korkas, K. K., & Fryzlewicz, P. (2017). Multiple change-point detection for non-stationary time series using wild binary segmentation. Statistica Sinica, 27(1), 287–311. doi:10.5705/ss.202015.0262.
[28] Wang, Y., Wang, J., Zhang, H., & Song, J. (2025). Bridging prediction and decision: Advances and challenges in data-driven optimization. Nexus, 2(1), 100057. doi:10.1016/j.ynexs.2025.100057.
[29] Li, D., Yang, J., Kreis, K., Torralba, A., & Fidler, S. (2021). Semantic Segmentation with Generative Models: Semi-Supervised Learning and Strong Out-of-Domain Generalization. 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 8296–8307. doi:10.1109/cvpr46437.2021.00820.
[30] Doan, Q. H., Mai, S.-H., Do, Q. T., & Thai, D.-K. (2022). A cluster-based data splitting method for small sample and class imbalance problems in impact damage classification. Applied Soft Computing, 120, 108628. doi:10.1016/j.asoc.2022.108628.
[31] Stevanović, S., Dashti, H., Milošević, M., Al-Yakoob, S., & Stevanović, D. (2024). Comparison of ANN and XGBoost surrogate models trained on small numbers of building energy simulations. PLoS ONE, 19(10), 312573. doi:10.1371/journal.pone.0312573.
[32] Mohammed, A., & Kora, R. (2023). A comprehensive review on ensemble deep learning: Opportunities and challenges. Journal of King Saud University - Computer and Information Sciences, 35(2), 757–774. doi:10.1016/j.jksuci.2023.01.014.
[33] Zhong, Y., Ren, Y., Cao, G., Li, F., & Qi, H. (2025). Optimal starting point for time series forecasting. Expert Systems with Applications, 273, 126798. doi:10.1016/j.eswa.2025.126798.
[34] Lebedev, I. S., & Sukhoparov, M. E. (2024). Improving the Quality Indicators of Multilevel Data Sampling Processing Models Based on Unsupervised Clustering. Emerging Science Journal, 8(1), 355–371. doi:10.28991/ESJ-2024-08-01-025.
[35] Nguyen, D. D., Tiep, N. V., Bui, Q. A. T., Van Le, H., Prakash, I., Costache, R., Pandey, M., & Pham, B. T. (2025). Landslide Susceptibility Mapping Using RBFN-Based Ensemble Machine Learning Models. CMES - Computer Modeling in Engineering & Sciences, 142(1), 467–500. doi:10.32604/cmes.2024.056576.
[36] Xu, S., Liu, J., Huang, X., Li, C., Chen, Z., & Tai, Y. (2024). Minutely multi-step irradiance forecasting based on all-sky images using LSTM-InformerStack hybrid model with dual feature enhancement. Renewable Energy, 224, 120135. doi:10.1016/j.renene.2024.120135.
[37] Wang, C., Li, X., Zhou, T., & Cai, Z. (2024). Unsupervised Time Series Segmentation: A Survey on Recent Advances. Computers, Materials & Continua, 80(2), 2657–2673. doi:10.32604/cmc.2024.054061.
[38] Kerr, D. J., & Yang, L. (2024). Viewpoint: prognostic markers can personalize adjuvant therapeutic options. Academia Oncology, 1(2). doi:10.20935/acadonco7393.
[39] Lebedev, I. S., & Sukhoparov, M. E. (2023). Adaptive Learning and Integrated Use of Information Flow Forecasting Methods. Emerging Science Journal, 7(3), 704–723. doi:10.28991/ESJ-2023-07-03-03.
[40] Liu, C., Kowal, D. R., Doss-Gollin, J., & Vannucci, M. (2025). Bayesian functional graphical models with change-point detection. Computational Statistics & Data Analysis, 206, 108122. doi:10.1016/j.csda.2024.108122.
[41] Osipov, V., Nikiforov, V., Zhukova, N., & Miloserdov, D. (2020). Urban traffic flows forecasting by recurrent neural networks with spiral structures of layers. Neural Computing and Applications, 32(18), 14885–14897. doi:10.1007/s00521-020-04843-5.
[42] Semenov, V. K. (2025). Sequence-segmentation. GitHub, San Francisco, United States. Available online: https://github.com/vksemenov/sequence-segmentation (accessed on September 2025).
[43] Lu, K. P., & Chang, S. T. (2023). An Advanced Segmentation Approach to Piecewise Regression Models. Mathematics, 11(24), 4959. doi:10.3390/math11244959.
[44] Xhu, X. (2010). Stream Data Mining Repository. Florida Atlantic University, Boca Raton, United States. Available online: http://www.cse.fau.edu/~xqzhu/stream.html (accessed on September 2025).
[45] MathWorks (2025). Regression, The MathWorks, Natick, United States. Available online: https://www.mathworks.com/help/stats/regression-and-anova.html?s_tid=CRUX_lftnav (accessed on September 2025).
[46] Kaggle. (2017). E-Commerce Data. Kaggle, Mountain View, United States. Available online: https://www.kaggle.com/datasets/carrie1/ecommerce-data (accessed on September 2025).
[47] Kaggle. (2019). Hourly energy demand generation and weather. Kaggle, Mountain View, United States. Available online: https://www.kaggle.com/nicholasjhana/energy-consumption-generation-prices-and-weather/data?select=energy_dataset.csv (accessed on September 2025).
[48] Kaggle. (2016). Pima Indians Diabetes Database. Kaggle, Mountain View, United States. Available online: https://www.kaggle.com/datasets/uciml/pima-indians-diabetes-database (accessed on September 2025).
[49] Piernik, M., & Morzy, T. (2021). A study on using data clustering for feature extraction to improve the quality of classification. Knowledge and Information Systems, 63(7), 1771–1805. doi:10.1007/s10115-021-01572-6.
[50] Etemad, A., Grover, S., & Jalali, A. (2024). Segment, Shuffle, and Stitch: A Simple Layer for Improving Time-Series Representations. Advances in Neural Information Processing Systems 37, 4878–4905. doi:10.52202/079017-0158.
[51] Shmuel, A., Glickman, O., & Lazebnik, T. (2024). Symbolic regression as a feature engineering method for machine and deep learning regression tasks. Machine Learning: Science and Technology, 5(2), 25065. doi:10.1088/2632-2153/ad513a.
[52] Otelbaev, M., Durmagambetov, A. A., & Seitkulov, Y. N. (2008). Conditions for existence of a global strong solution to one class of nonlinear evolution equations in Hilbert space. Siberian Mathematical Journal, 49(3), 498–511. doi:10.1007/s11202-008-0048-2.
[53] Petukhov, I., Steshina, L., & Tanryverdiev, I. (2014). Remote Sensing of Forest Stand Parameters for Automated Selection of Trees in Real-Time Mode in the Process of Selective Cutting. 2014 IEEE 11th Intl Conf on Ubiquitous Intelligence and Computing and 2014 IEEE 11th Intl Conf on Autonomic and Trusted Computing and 2014 IEEE 14th Intl Conf on Scalable Computing and Communications and Its Associated Workshops, 390–395. doi:10.1109/uic-atc-scalcom.2014.113.
- This work (including HTML and PDF Files) is licensed under a Creative Commons Attribution 4.0 International License.
