DocumentCode :
3130778
Title :
Analog implementation of the incremental credit assignment (ICRA) scheme for time series classification
Author :
Hatzopoulos, A. ; Koukas, G. ; Mpinos, G. ; Petridis, V. ; Kehagias, A.
Author_Institution :
Dept. of Electr. & Comput. Eng., Aristotelian Univ. of Thessaloniki, Greece
Volume :
2
fYear :
1997
fDate :
2-4 Jul 1997
Firstpage :
731
Abstract :
The incremental credit assignment (ICRA) scheme has been introduced by Petridis et al. (1996) and its application to time series classification has been shown by simulation results. In this work an analog hardware implementation of its decision module is described and its practical characteristics are given. The performance of the analog circuit agreed with the theoretical simulations and the results obtained were very promising for the application of the circuit in practical classification problems, like speech recognition
Keywords :
analogue multipliers; feedforward neural nets; pattern classification; time series; decision module; feedforward neural networks; incremental credit assignment; multipliers; speech recognition; time series classification; Adders; Analog computers; Circuit simulation; Computer networks; Computer simulation; Concurrent computing; Convergence; Hardware; Neural networks; Neurons;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Digital Signal Processing Proceedings, 1997. DSP 97., 1997 13th International Conference on
Conference_Location :
Santorini
Print_ISBN :
0-7803-4137-6
Type :
conf
DOI :
10.1109/ICDSP.1997.628456
Filename :
628456
Link To Document :
بازگشت