Title :
Behavior discovery as database scheme design
Author :
Miura, Takao ; Shioya, Isamu ; Watanabe, Kohei
Author_Institution :
Dept. of Electron. & Electr. Eng., Hosei Univ., Koganei, Japan
Abstract :
In this investigation, we propose an inductive classification method for a collection of temporal information which mean behavior. We assume a Markov property and discuss ergodic analysis by which we can extract stationary parts of the behavior. We discuss a design methodology for database schemes based on the theory
Keywords :
data mining; database theory; learning (artificial intelligence); temporal databases; Markov property; data mining; database scheme design; design methodology; ergodic analysis; inductive classification method; learning; temporal database; temporal learning; Databases;
Conference_Titel :
Temporal Representation and Reasoning, 2000. TIME 2000. Proceedings. Seventh International Workshop on
Conference_Location :
Cape Breton, NS
Print_ISBN :
0-7695-0756-5
DOI :
10.1109/TIME.2000.856592
