中文说明：ELM算法是由南洋理工大学黄广斌副教授提出，相比于BP算法复杂的参数设置，该算法结构简单，只需输入隐层节点数即可，输入权值和阈值随机输入成后为固定值，输出权值由隐含层输出与网络输出数据确定，因此ELM算法属于前馈型神经网络，并且学习速度较快、泛化能力好。
English Description：
Elm algorithm is proposed by Associate Professor Huang guangbin of Nanyang University of technology. Compared with the complex parameter setting of BP algorithm, the structure of elm algorithm is simple. It only needs to input the number of hidden layer nodes, and the input weight and threshold are fixed after random input. The output weight is determined by the hidden layer output and network output data. Therefore, elm algorithm belongs to feedforward neural network, and its learning speed is fast Good generalization ability.