基于协同过滤的滑坡治理工程方案研究

Research on Landslide Treatment Engineering Scheme Based on Collaborative Filtering

以滑坡为研究对象,结合机器学习中基于用户的协同过滤推荐算法,通过对已发生的滑坡及其治理措施进行数据分析和特征提取,建立预测模型。对于一个新发生未治理的滑坡,计算其与现有所有已治理滑坡的相似度,统计与目标滑坡相似度最高的前K个滑坡的治理措施,最快地给出治理工程方案建议,能够有效地降低滑坡的破坏性并减少损失。

This article takes landslide as research object,combining user-based collaborative filtering recommendation algorithms in machine learning,build predictive models based on data analysis and feature extraction of existing landslides and their governance measures.For a new landslide,calculate the similarity with all existing managed landslides,statistics of the top K landslide governance measures with the highest similarity to the target landslide,and give the fastest suggestion for the treatment project which can effectively reduce the destructiveness of landslides and reduce losses.