Benchmarks for testing evolutionary algorithms
Wenying Feng, T. Brune, Lipton Chan, Mosharaf Chowdhury, C.K. Kuek, Yang Li · ENLIGHTEN (Jurnal Bimbingan dan Konseling Islam) · 1998
Although some convergence proofs of evolutionary and other optimisation algorithms exist, most global convergence rates and qualities can only be measured or compared through benchmark tests. Benchmarks that may be used to assess the merit and performance of these algorithms should be simple to use, widely applicable and reliable. This paper attempts to address, and to promote discussions on, this issue. In particular, benchmarks in term of (i) optimality;(ii) accuracy;(iii) sensitivity;(iv) convergence and (v) ...