The authors present a modified mean field annealing algorithm which supports the combinatorial optimisation problems with continuous real-valued states. The proposed algorithm has been applied to the construction of D-optimal designs for a simple polynomial regression model with degree 5. Experimental results show that the proposed algorithm is useful and effective in terms of computation time and the quality of solutions, compared with the stochastic simulated annealing algorithm.
References
-
-
1)
-
D.E. Van den Bout ,
T. Miller
.
Graph partitioning using annealed neural networks.
IEEE Trans. Neural Networks
,
192 -
203
-
2)
-
L.M. Haines
.
The application of the annealing algorithm to the construction of exactoptimal designs for linear-regression models.
Technometrics
,
439 -
447
-
3)
-
N. Ansari ,
E.S.H. Hou ,
Y. Yu
.
A new method to optimise the satellite broadcasting schedules using themean field annealing of a Hopfield neural network.
IEEE Trans. Neural Networks
,
470 -
483
-
4)
-
G. Bilbro ,
R. Mann ,
T. Miller ,
W. Snyder ,
D.E. Van den Bout ,
M. White
.
Optimisation by mean field annealing.
Adv. Neural Info. Process. Syst.
,
91 -
98
-
5)
-
S. Kirkpatrick ,
C.D. Gelatt ,
M.P. Vecchi
.
Optimisation by simulated annealing.
Science
,
671 -
680
-
6)
-
C. Peterson ,
J.R. Anderson
.
A mean field theory learning algorithm for neural networks.
Complex Systems
,
995 -
1019
http://iet.metastore.ingenta.com/content/journals/10.1049/el_19970641
Related content
content/journals/10.1049/el_19970641
pub_keyword,iet_inspecKeyword,pub_concept
6
6