Detecting Regime Change in Financial Data with an Auto-Encoder
Refereed conference paper presented and published in conference proceedings


Full Text

Other information
AbstractFinancial markets are dynamic, not just on the level of observed prices and volumes, but also on the level of underlying logic. Unlike physical systems, the underlying distribution may change suddenly and dramatically. The market may settle in one regime for a period of time, then suddenly switch to a different regime, but then switch back in the next period. When such regime changes happen, investors often experience very different market situations. Old and successful investment strategies fail, and market participation logic needs to be revised.
All Author(s) ListLe Zhang, Tao Lu, Michael Zhang, Weidong Liu
Name of ConferenceStatistical Challenges in E-Commerce Research (SCECR)
Start Date of Conference10/06/2019
End Date of Conference13/06/2019
Place of ConferenceHong Kong
Country/Region of ConferenceChina
Year2019
Month6
LanguagesEnglish-United States

Last updated on 2021-08-01 at 14:27