Swarm Intelligence Algorithms

Conclusion 119

Once the above modifications over the whole population and updates are

done, we move on to the next generation t = 2 of iterations.

Cuckoo search is very efficient, even with a small population. For example,

if we use n = 5 (cuckoos), we can get f

best

= 1.34×10

−17

after 1000 iterations

for the above example. Even for D = 15 and n = 25, we can typically get

f

best

= 2.50 × 10

−12

after 1000 iterations.

Obviously, due to the stochastic nature, performance evaluations should

be based on the statistical measures of results of multiple runs with proper

parametric studies.

9.6 Conclusion

The cuckoo search algorithm can be very efficient and flexible. Ho