報(bào)告人：孔令臣教授，北京交通大學(xué)數(shù)學(xué)與統(tǒng)計(jì)學(xué)院

報(bào)告題目：Fast Online Elastic Net Subspace Clustering via a Novel Dictionary Update Strategy

報(bào)告時(shí)間：2024年12月13日，星期五，下午：13:30-14:30

報(bào)告地點(diǎn)：騰訊會(huì)議號(hào)：447 4167 6172

報(bào)告摘要：In recent years, online subspace clustering has emerged as a critical tool for real-time analysis of data streams in various applications such as video surveillance and social media analytics. While traditional subspace clustering methods may not adequately capture the complex and varying structures in dynamically changing data streams. In this paper, a fast online elastic net subspace clustering model with block diagonal property is introduced, which can be adapted to varying data characteristics while maintaining robustness against noise and outliers. Furthermore, the dynamic nature of online data requires that the model update its parameters efficiently without reprocessing the entire dataset. To meet this need, an alternating direction method of multipliers method with a novel dictionary update strategy based on support point is designed. This dictionary update strategy not only enhances the adaptability of the model by selectively updating the dictionary atoms that best represent the current data characteristics but also significantly enhances the computational efficiency. Moreover, we rigorously prove the convergence of the algorithm, thereby ensuring its reliability and stability in practical applications. Finally, extensive numerical experiments demonstrate that the proposed method not only improves the accuracy of subspace clustering, but also maintains scalability, making it suitable for real-time and large-scale data processing tasks.

報(bào)告人簡(jiǎn)介：孔令臣，教授，博士生導(dǎo)師，中國(guó)運(yùn)籌學(xué)會(huì)數(shù)學(xué)規(guī)劃分會(huì)理事長(zhǎng)，北京交通大學(xué)數(shù)學(xué)與統(tǒng)計(jì)學(xué)院副院長(zhǎng)。主要從事對(duì)稱錐互補(bǔ)問(wèn)題和最優(yōu)化、高維數(shù)據(jù)分析、統(tǒng)計(jì)優(yōu)化與學(xué)習(xí)、醫(yī)學(xué)成像等方面的研究。在《Mathematical Programming》《SIAM Journal on Optimization》《IEEE Transactions on Pattern Analysis and Machine Intelligence》《IEEE Transactions on Signal Processing》《Technometrics》《Statistica Sinica》《Electronic Journal of Statistics》等期刊發(fā)表論文60余篇。主持國(guó)家自然科學(xué)基金面上項(xiàng)目和專項(xiàng)基金項(xiàng)目, 參與國(guó)家自然科學(xué)基金重點(diǎn)項(xiàng)目、重點(diǎn)研發(fā)項(xiàng)目以及973課題等。2012年獲中國(guó)運(yùn)籌學(xué)會(huì)青年獎(jiǎng)，2018年獲得北京市高等教育教學(xué)成果一等獎(jiǎng)，2022年獲教育部自然科學(xué)二等獎(jiǎng)等。