报告题目：Interactions and information transfer across time scales
报告时间：2019年10月17日下午2：00
报告地点：文德楼S517
报 告 人：Dr. Milan Paluš
主 持 人：梁湘三 教授
摘要：Information-theoretic generalization of Granger causality in the form of conditional mutual information, also known as transfer entropy is widely used in various areas of science, including the study of complex systems. Complex systems typically evolve on multiple time scales. We will describe how to analyse dynamics on different scales and demonstrate information transfer across time scales in climate dynamics, namely in the air temperature variability in European mid-latitudes and in the El Niňo Southern Oscillation. Then we will discuss some issues related to time i/reversibility in causality analysis.
专家简介：Dr. Milan Paluš，现为捷克科学院计算机研究所的高级研究员，此前在Prague Psychiatric Centre从事确定性混沌在脑电波分析中的应用研究。Paluš教授对复杂系统、非线性动力学、确定性混沌理论以及因果分析均有重要贡献，一直致力于开发检测和表征（多变量）时间序列之间同步现象的方法。
海洋科学学院
2019年10月15日