为了评价ELM的性能，试分别将ELM应用于基于近红外光谱的汽油辛烷值测定和鸢尾花种类识别两个问题中,并将其结果与传统前馈网络(BP、RBF、PNN、GRNN等)的性能和运行速度进行比较,并探讨隐含层神经元个数对ELM性能的影响。(n order to evaluate the performance of ELM, test the application of ELM in the near infrared spectroscopy based on the determination of gasoline octane value and two kinds of problems of iris recognition, and the results of the traditional feedforward networks (BP, RBF, PNN, GRNN) to compare the performance and speed, and to explore the hidden layer number effect on the performance of ELM neurons.)