1315 IEEE TRANSACTIONS ON NEURAL NETWORKS, VOL. 7, NO. 5 SEPTEMBER 1996
Extemal
reinforcementsignal
Corrections to "Reinforcement Learning for an
ART-Based Fuzzy Adaptive Learning Control Network"
Cheng-Jian Lin and Chin-Teng Linqtemal
'
-
rein orcement In the above paper,' the graphics for Figs. 3,5,
and6
were printedincorrectly due to a publisher error. The correct figures an:
as
follows.r ( t + 1)
'I
P O +
1)'
_ _ _ - - -
Predicted y24
signal
jl
( t )
s
(
t
+
1)
Action
21"22 IH3 OH2 where
'23
Critic
Network
Fig. 5. The fuzzy reasoning process in the FALCON model.
Action
Network
Input
States
x
( t )
Fig. 6. The proposed RFALCON.
Fig. 3 . The proposed FALCON. Manuscript received July 1, 1996.
The authors are with the Department of Control Engineering, National
Publisher Item Identifier S 1045-9227(96)08055-1.
'
C.-J. Lin and C.-T. Lin, IEEE Trans. Neural Networks, vol. ' 7 , pp. 709-73 1, Chiao-Tung University, Hsinchu, Taiwan, R.O.C.May 1996.