嘉宾简介：

Wolfgang Karl Härdle教授是德国柏林洪堡大学Ladislaus von Bortkiewicz讲席教授。他出版专著40余部，Google Scholar引用量位居德国柏林洪堡大学首位，H指数达94，是数学、经济学与金融学领域的领军学者。他是BRC区块链研究中心的创始人，创立了布加勒斯特经济研究学院数字资产研究所（IDA），并指导了中德国际研究训练项目IRTG1792 。他发明了CRIX加密货币指数，担任多所亚洲高校的高级顾问，在机器学习与可解释人工智能领域有很多卓越贡献。

讲座简介：

A radar-based framework is investigated that allows for reliable discrimination between unmanned aerial vehicles (UAVs) and birds under low-visibility conditions. Using complex-valued mmWave MIMO radar data, targets are localized via phase-coherence based region-of-interest selection and decomposed by complex empirical mode decomposition (CEMD). Physically interpretable micro-Doppler features—instantaneous frequency, bandwidth, energy, and spectral entropy of intrinsic mode functions—are extracted and embedded in a supervised classification pipeline combining PCA principal component analysis and SVMs support vector machines. Experiments on controlled anechoic-chamber measurements of drones and a bionic bird demonstrate stable classification accuracies around 95% and robustness under additive noise, highlighting the suitability of the approach for counter-UAS applications.

讲座提醒：本讲座的ppt和程序，可通过网页quantinar.com免费获取。建议大家带 laptop 参加讲座，可以更好地体验相关的分析工作。