基于随机森林算法的张家界生态旅游适宜性评价研究  

Evaluation of Zhangjiajie Ecotourism Suitability Based on Random Forest Algorithm

作　　者：杨波[1] 黄钦 郑群明[2] 龚熊波 梁莉莉 王敏[1] 陈颖[1] 袁慧芳 YANG Bo;HUANG Qin;ZHENG Qun-ming;GONG Xiong-bo;LIANG Li-li;WANG Min;CHEN Ying;YUAN Hui-fang(Hunan Key Laboratory of Geospatial Big Data Mining and Application,College of Geographic Science,Hunan Normal University,Changsha 410081,China;College of Tourism,Hunan Normal University,Changsha 410081,China)

机构地区：[1]湖南师范大学地理科学学院,地理空间大数据挖掘与应用湖南省重点实验室,中国长沙410081 [2]湖南师范大学旅游学院,中国长沙410081

出　　处：《湖南师范大学自然科学学报》2021年第4期17-25,共9页Journal of Natural Science of Hunan Normal University

基　　金：国家自然科学基金资助项目(41171342,41071067);湖南省教育厅重点项目(17A127);湖南省哲学与社会科学基金资助项目(16JD48)。

摘　　要：生态旅游适宜性评价是进行旅游资源开发的前提和基础,是制定旅游空间规划的科学依据,以生态旅游为核心的适宜性评价研究对于生态保护与旅游开发意义重大。本文在参考前人方法的基础上,引入机器学习方法,通过对小规模已知样本的训练学习实现大规模未知数据的高精度分类与评价,从方法科学性与可行性、因子选取、数模结合、预测实现、精度检验等方面展开探讨,并运用随机森林算法对张家界生态旅游适宜性进行实证研究。研究表明,随机森林算法能够较好地支撑生态敏感区的旅游适宜性评价工作,模型优化后的测试精度为94.20%,其评价结果能够为优化生态旅游景区空间规划提供科学依据。Ecotourism suitability evaluation is the prerequisite for the development of tourism resources and the scientific basis for the development of tourism spatial planning.The suitability evaluation research centered on ecotourism is of great significance to ecological protection and tourism development.Based on previous methods reported in the literature,this paper introduces machine learning methods,and achieves high-precision classification and evaluation of large-scale unknown data through training and learning of small-scale known samples.Factors such as the feasibility of methods,selection of factors,and model combination,prediction realization,accuracy test and others were taken into consideration.The random forest algorithm was applied to conduct empirical research on the suitability of ecological tourism in Zhangjiajie.Our results show that the random forest algorithm can better support the evaluation of tourism suitability in ecologically sensitive areas.The test accuracy of the optimized model is 94.20%.Our evaluation results can provide the scientific basis to optimize the spatial planning of eco-tourism attractions.

关 键 词：机器学习 随机森林算法 生态旅游 适宜性评价 张家界 

分 类 号：F592[经济管理—旅游管理]