中文摘要：

运用烟花算法（fireworks algorithm,FWA）优化极限学习机（extreme learning machine,ELM）.首先烟花算法经过多次的迭代,确定M个最优的烟花,并且以极限学习机测试样本的RMSE作为烟花算法每次迭代的适应度函数,达到优化极限学习机的输入权值矩阵和隐含层偏差的效果.最后根据广义逆求出输出矩阵.通过对一维sin C函数的测试结果表明,烟花算法优化极限学习机能够以较少的隐含层节点数目达到更高的精度,比极限学习机的测试误差降低了29.58%.在以上基础上又做了对高斯正态分布函数的拟合实验,验证了烟花算法优化极限学习机比极限学习机拥有更好的拟合性能.

英文摘要：

The fireworks algorithm（ FWA） is used to optimize the extreme learning machine（ ELM） in this paper. Firstly,the FWA gain the M optimal fireworks through many iterations,and the RMSE of the extreme learning machine＇s test samples is used as the fitness function in each iteration. Secondly,the optimization of the input weights and hidden layer deviation matrix of the extreme learning machine is achieved. Finally,the matrix output is obtained based on the generalized inverse. The test experiment of one-dimensional sin C function is conducted. The experimental results show that the fireworks algorithm extreme learning machine achieves higher accuracy with less number of hidden layer nodes,and the test error decreases 29. 58% compared with the extreme learning machine. The fitting experiment of Gauss normal distribution function is conducted,and the experiment results further demonstrate that the FWAELM achieves a better fitting effect than the ELM.