基于优化核极限学习机的泥石流危险性评估

山区环境中泥石流的孕育受多种因素的影响，为提高泥石流危险性的预测精度，提出一种萤火虫算法(firefly algorithm, FA)优化核极限学习机(kernel based extreme learning machine, KELM)的预测模型。首先，针对数据维度爆炸的问题，通过主成分分析(principal component analysis, PCA)数据降维，使得留有大部分致灾特征信息的因子输入训练模型；然后，使用萤火虫优化算法更新核极限学习机的参数，将四川省北川县监测数据输入优化后的预测模型，并与其他传统机器学习算法进行对比分析，验证该算法的优越性；最后，使用多种指标综合评估模型的预测效果。结果表明，FA-KELM模型能够有效地简化数据结构，提高泥石流危险性预测的准确性，为泥石流灾害预测方面的研究提供参考和借鉴。

Debris Flow Hazard Assessment Based on Optimized Kernel Extreme Learning Machine

The breeding of debris flow is affected by many factors in mountainous environment . In order to improve the prediction accuracy of debris flow risk, this paper proposed a prediction model that firefly algorithm (FA) to optimize kernel based extreme learning machine (KELM). Firstly, to solve the problem of data dimension explosion, the principal component analysis (PCA) was used to reduce the dimension of data, so that the factor with most of the feature information was input into the training model. Then, the firefly algorithm was used to update the parameters of the kernel based extreme learning machine, the monitoring data of Beichuan County, Sichuan Province were input into the optimized prediction model.The FA-KELM model compared with other traditional machine learning algorithms to verify the superiority of the algorithm. Finally, a variety of indicators were used to evaluate the prediction effect of the model. The results show that the FA-KELM model can effectively simplify the data structure, improve the accuracy of debris flow hazard prediction, and provide a reference for the research of debris flow disaster prediction.