The purpose of this paper is to find the event memory characteristics hidden in trade data.
First, historical trade data are analyzed to define the events described by multi‐dimensional characteristic variables. The variables containing information are employed to build the event description patterns. Furthermore, a search engine is developed for calendar events, which can search for events in historical data and produce a collection of events. The search engine also extracts relevant system reaction phenomena described by trend distribution for each event pattern. Finally, both event patterns and system reactions construct the episodic memory model.
The event patterns and the system reactions are used to define the episodic memory model. The search methods for the episodic memory model obtained from trade data set are given.
Accessibility and availability of data are the main limitations affecting where the method can be applied.
The method is helpful for traders when judging the current trade situation from historical memory.
The paper presents a new episodic memory modeling method based on trade data.
