Hierarchical Segmentation Approach to Detecting Switching Interaction using Bayesian Non-Parametrics
Briones Jeric Cruz · Institutional Repositories DataBase (IRDB) · 2020
1、、,'O C0皿mon topics in time series analysis are seg皿entation and interaction detection.On one hand, existing se晉mentation methods either estimate the dynamics ofthe ti血e Series, or extract 此S hierarchical structure, but usua11y not both.However, we are interested in a se套mentation method that discovers both the dynamics and structure of the data.on tbe other hand, existin目 interaction detection methodS 套enera11y assume that interaction information is tim.e'invariant.However, we are inteTested in detectin晉 interactionthat chan套es overtime, since dynamics oftime seTies sequences may sW北Ch.While these t剖'o problems aぞe seemingly unconnected, we want to use a se号mentation