基于不确定的数据挖掘分类算法在滑坡灾害预测的应用

发布时间：2018-04-20 11:05

本文选题：数据挖掘 + 决策树　；参考：《江西理工大学》2015年硕士论文

【摘要】：中国是地质灾害的多发国家,尤其是滑坡灾害,一旦发生降雨,一些地方就很容易发生滑坡,对人们的生命造成危害,对财产造成损失。因此,如何高效的对区域滑坡灾害进行预防,是一个很重要的课题。数据挖掘是一种新兴的数据分析技术,它能通过分类方法进行学习,提取出规则,从而对未知的事物进行预测。而滑坡灾害预测受多种因素影响,其中降雨等不确定因素存在难以获取数据及有效处理等难题,为提高滑坡危险性预测的准确率,根据滑坡灾害发生相关理论及数据挖掘中决策树分类原理,本文提出了两种处理方法:基于不确定的C4.5决策树算法和不确定模糊ID3决策树算法,分别对滑坡危险性进行预测。本文首先进行了滑坡灾害和数据挖掘相关的理论介绍,为后面的章节打下理论基础。然后介绍了精度较高的不确定C4.5决策树算法,并将它应用于实例中,进行效果检验。接着在传统模糊ID3算法的基础上,利用积分思想,设计出不确定数据的模糊化处理方法,并结合模糊ID3算法,提出了一种新的不确定模糊ID3算法,建立不确定模糊ID3决策树模型,对区域滑坡危险性进行预测,在简化算法复杂度的基础上,也保证了算法的精确度。最后,将传统的算法和本文提出的算法进行比较,分析两种算法模型的优缺点,推断出各自的适用场合,为以后该类问题的算法选择提供参考。
[Abstract]:China is a country prone to geological disasters, especially landslide disasters. Once rainfall occurs, some places are prone to landslides, causing harm to people's lives and property losses. Therefore, how to efficiently prevent the regional landslide disaster is a very important subject. Data mining is a new technology of data analysis. It can learn from classification method, extract rules and predict unknown things. However, landslide disaster prediction is affected by many factors, such as rainfall and other uncertain factors, which are difficult to obtain data and deal with effectively, in order to improve the accuracy of landslide hazard prediction. According to the related theory of landslide occurrence and the classification principle of decision tree in data mining, this paper proposes two processing methods: C4.5 decision tree algorithm based on uncertainty and fuzzy ID3 decision tree algorithm based on uncertainty, respectively, to predict landslide risk. This paper firstly introduces the theory of landslide disaster and data mining, which lays a theoretical foundation for the later chapters. Then, the uncertain C4.5 decision tree algorithm with high precision is introduced, and it is applied to an example to verify the effect. On the basis of the traditional fuzzy ID3 algorithm, the fuzzy processing method of uncertain data is designed by using integral idea, and a new uncertain fuzzy ID3 algorithm is proposed based on fuzzy ID3 algorithm. An uncertain fuzzy ID3 decision tree model is established to predict the landslide risk in the region. Based on the simplification of the complexity of the algorithm, the accuracy of the algorithm is ensured. Finally, the traditional algorithm and the algorithm proposed in this paper are compared, the advantages and disadvantages of the two models are analyzed, and their respective applicable situations are inferred, which provides a reference for the selection of the algorithm for this kind of problem in the future.
【学位授予单位】：江西理工大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P642.22;TP311.13

【参考文献】