活動日期: 2024/04/22(一)
地點:研究大樓250203
承辦單位:應數系
聯絡人:陸芷昀 0229393091分機62370
聯絡人:陸芷昀 0229393091分機62370
E-mail:lu1117@nccu.edu.tw
講座姓名:李智新
講座現職:聖路易斯華盛頓大學講座教授
講題:舞動機器、智慧生命:動態系統學習之美
探索機器學習的新視野!
你聽過動態系統嗎?
從大腦醫學到生物網絡,動態系統的概念被用於促進先進的學習理論和算法。
當機器學習遇到動態系統,又會碰撞出什麼新火花?
逆向工程,打破框架, 啟發未來的學習模式!
千萬不要錯過這精彩的驚奇之旅!
【Abstract】
Machine learning has become a prominent and mainstream area of research in the past decade. More
recently, there has been a significant surge of interest in learning dynamical systems from their measurement
data. This is attributed to its fundamental importance in various cross-cutting applications from
brain medicine for understanding brain dynamics to chronobiology for mapping functional connectivity
in large-scale biological networks. On the other side of the spectrum, there is a growing concept
of exploiting the notions and tools of dynamical systems to facilitate advanced learning theory and algorithms.
These approaches go beyond the limitations of conventional machine learning methods and
catalize the emerging idea of inverse engineering. The interconnected concepts of learning for dynamical
systems and dynamical systems for learning form the cornerstone for constructing the future building
blocks of machine learning. This talk will address theoretical and computational approaches to learning
for dynamical systems, and illustrate systems-enabled learning theory and data-driven algorithms.
These advancements will be demonstrated through various examples that have substantial practical implications.
Some notable examples include geometric embedding of dynamic data, network inference,
topological manifold learning, and pattern formation and control in complex ensemble systems enabled
by optimal transport.
<<本活動報名確認者提供餐點>>
<<本場次為中文演講>>