تحلیل تطبیقی عملکرد سیستمهای دستهبند یادگیر قابل تبیین با استفاده از دادههای جریان و الگوریتمهای تکامل پویا
محورهای موضوعی : مهندسی کامپیوتر و فناوری اطلاعات
محمدرضا دهقانی محمودآبادی
1
*
,
الهام دهقان طزرجانی
2
1 - گروه مهندسی کامپیوتر، دانشکده فنی و مهندسی، دانشگاه آزاد اسلامی واحد بافق، بافق، ایران.
2 - گروه مهندسی کامپیوتر، دانشکده فنی و مهندسی، دانشگاه آزاد اسلامی واحد بافق، بافق، ایران.
کلید واژه: سیستمهای دستهبند یادگیر, جریان داده, تغییر مفهوم, الگوریتمهای تکاملی پویا, تبیینپذیری.,
چکیده مقاله :
با رشد فزاینده دادههای جریانی در سامانههای هوشمند، نیاز به روشهای یادگیری تطبیقی و قابلتبیین بیشازپیش احساس میشود. سیستمهای دستهبندِ یادگیر، با تلفیق یادگیری تقویتی و جستوجوی تکاملی، بستر مناسبی برای تحلیل دادههای پویا فراهم میکنند. در این پژوهش، عملکرد نسخههای پیشرفته و قابلتبیین سیستمهای دستهبندِ یادگیر در مواجهه با دادههای جریانی، با استفاده از الگوریتمهای تکامل پویا مورد بررسی قرار گرفته است. بدین منظور، مکانیزمهایی برای شناسایی و انطباق با تغییرِ مفهوم طراحی و پیادهسازی شدهاند. سپس، با بهرهگیری از مجموعهدادههای جریانی استاندارد در حوزههای متنوع، از جمله امنیت شبکه و پایش حسگرها، تحلیل تطبیقی میان مدلهای پیشنهادی و سایر روشهای دستهبندی جریانی، مانند درختِ هوفدینگ و جنگلِ تصادفیِ تطبیقی، انجام گرفت. نتایج نشان میدهد که مدل پیشنهادی، ضمن حفظ تبیینپذیریِ قوانین، توانایی تطبیق سریع با تغییرات محیطی و حفظ دقتِ دستهبندی در شرایط پویای داده را داراست. این مطالعه گامی مؤثر در جهت توسعه سامانههای تصمیمیارِ بلادرنگ و شفاف در محیطهای دادهمحور بهشمار میآید.
With the exponential growth of streaming data in intelligent systems, the demand for adaptive and explainable learning methods has become more critical than ever. Learning Classifier Systems, which integrate reinforcement learning with evolutionary search, provide a robust foundation for analyzing dynamic data streams. This study investigates the performance of advanced, explainable Learning Classifier System variants in handling streaming data by leveraging dynamic evolutionary algorithms. Novel mechanisms are designed and implemented to effectively detect and adapt to concept drift. The proposed models are evaluated on standard benchmark streaming datasets across diverse domains, including network security and sensor monitoring, and are compared against state-of-the-art stream classification methods such as the Hoeffding Tree and Adaptive Random Forest. Experimental results reveal that the proposed models preserve rule interpretability while achieving rapid adaptation to environmental changes and maintaining high classification accuracy in dynamic scenarios. This work marks a significant step toward building transparent, real-time decision-support systems in data-driven environments.
[1] R. G. N. Paiva, Y. R. Melo, C. A.V. Cavalcante, V. A. S. Tenório, and P. Do, “Developing data-driven O&M policy through sequential pattern mining: A case study,” Comput Ind Eng, vol. 193, p. 110318, 2024.
[2] A. Zabala-López, M. Linares-Vásquez, S. Haiduc, and Y. Donoso, “A Survey of data-centric technologies supporting decision-making before deploying military assets,” Defence Technology, 2024.
[3] M. R. D. M. Abadi and H. Samavarchi, “The Effect of Fuzzy Logic on Database,” in 2009 WRI World Congress on Computer Science and Information Engineering, 2009.
[4] L. F. Niedermaier Custodio and C. de C. Carneiro, “Global enhanced oil recovery assessment: A comprehensive data-driven analysis using self-organizing maps,” Geoenergy Science and Engineering, vol. 237, p. 212771, 2024.
[5] S. Harno, H. Kai Chan, and M. Guo, “Enhancing value creation of operational management for small to medium manufacturer: A conceptual data-driven analytical system,” Comput Ind Eng, vol. 190, p. 110082, 2024.
[6] S. Subudhi and S. Panigrahi, “Application of OPTICS and ensemble learning for Database Intrusion Detection,” Journal of King Saud University - Computer and Information Sciences, vol. 34, no. 3, pp. 972–981, 2022.
[7] A. Zeb, J. Linnosmaa, M. Seppi, and O. Saarela, “Developing deep learning surrogate models for digital twins in mineral processing – A case study on data-driven multivariate multistep forecasting,” Miner Eng, vol. 216, p. 108867, 2024.
[8] R. Mohammedqasem et al., “Multi-objective deep learning framework for COVID-19 dataset problems,” J King Saud Univ Sci, vol. 35, no. 3, p. 102527, 2023.
[9] W. G. Krebs, V. Alexandrov, C. A. Wilson, N. Echols, H. Yu, and M. Gerstein, “Normal Mode Analysis of Macromolecular Motions in a Database Framework: Developing Mode Concentration as a Useful Classifying Statistic.,” Proteins, vol. 48, no. 4, pp. 682–695, 2002.
[10] Y. Hajjaji, W. Boulila, I. R. Farah, I. Romdhani, and A. Hussain, “Big data and IoT-based applications in smart environments: A systematic review,” Comput Sci Rev, vol. 39, p. 100318, 2021.
[11] S. Farshidi, B. Mohammadi, and H. Zarei, “Evolutionary Strategy Design for Mining,” Mining Strategy and AI Journal, 2022.
[12] N. Jain and P. K. Jana, “ L R F : A logically randomized forest algorithm for classification and regression problems ,” Expert Syst Appl, p. 119225, Mar. 2022.
[13] [N. Vural, Ö. Algan Cavuldak, and M. A. Akay, “D-Optimal design and multi-objective optimization for green extraction conditions developed with ultrasonic probe for oleuropein,” J Appl Res Med Aromat Plants, vol. 20, p. 100279, 2021.
[14] M. R. dehganimahmoudabadi N nabavizadeh, “Improving the learning classifier system with the basic memetic algorithm to solve the problem based on the law,” The journal of Engineering Management and Soft Computing., Feb. 2023.
[15] [15] Mohammadreza, “A novel aggregation-based distributed search model for optimal design using the memetic algorithm”.
[16] M. Dehghanimahmoudabadi, K. Mirzaie, and F. Peyravi, “Improving Credit Assignment in a learning classier system with Markov reinforcement learning for protein secondary structure prediction,” Journal of Applied and Basic Machine Intelligence Research, vol. 1, no. 2, pp. 92–104, 2023.
[17] M. DehghaniMahmoudAbadi and H. Samavarchi, “Designing the Matrix of Analysis to Solve the Problems Regarding the Web Site Searching,” International Journal of Soft Computing And Software Engineering (JSCSE), vol. 34, no. 2251–7545, p. DOI-10, 2013.
[18] J. Wang, L. Zhang, and Y. Liu, “AI-based Strategic Decision Making in Oil Sector,” Journal of Energy Strategy, 2020.
[19] R. Kumar, A. Singh, and P. Mehta, “Explainable AI in Pharma Strategy,” International Journal of Pharmaceutical Strategy, 2021.
[20] J. Casillas, B. Carse, and L. Bull, “Fuzzy-XCS: A Michigan Genetic Fuzzy System,” IEEE Transactions on Fuzzy Systems, vol. 15, no. 4, pp. 536–550, 2007.
[21] M. Nakata, T. Kovacs, and K. Takadama, “A Modified XCS Classifier System for Sequence Labeling,” 2014.
[22] P. L. Lanzi, “Using Raw Accuracy to Estimate Classifier Fitness in XCS,” 2003.
[23] A. R. Pakraei and K. Mirzaie, “The Introduction of a Heuristic Mutation Operator to Strengthen the Discovery Component of XCS,” Journal of Advances in Computer Research, vol. 9, no. 1, pp. 5170, 2018.
[24] J. E. Z. Macias and S. Trilles, “Machine learning-based prediction model for battery levels in IoT devices using meteorological variables,” Internet of Things, vol. 25, p. 101109, 2024.
[25] S. Maiti, A. Hassan, and P. Mitra, “Boosting Phosphorylation Site Prediction with Sequence Feature-Based Machine Learning.,” Proteins, vol. 88, no. 2, pp. 284–291, 2020.
[26] Masood, Z. Jiangbin, I. Ahmad, C. Dongdong, W. Shabbir, and M. Irfan, “A novel continual reinforcement learning-based expert system for self-optimization of soft real-time systems,” Expert Syst Appl, vol. 238, p. 122309, 2024.
[27] W. Wardah, M. G. M. Khan, A. Sharma, and M. A. Rashid, “Protein secondary structure prediction using neural networks and deep learning: A review,” Comput Biol Chem, vol. 81, pp. 1–8, 2019.
[28] M.-C. Chiu, Y.-J. Huang, and C.-J. Wei, “Enhancing SMEs Digital Transformation through Machine Learning: A Framework for Adaptive Quality Prediction,” J Ind Inf Integr, p. 100666, 2024.