论文部分内容阅读
本文以马铃薯甲虫为例,探索建立不同分析方法所预测结果之间的定量评估技术,并以此明确马铃薯甲虫在我国东北地区的适生范围和定殖风险。研究采用统一的环境数据、生物数据,分别利用GAM、Max Ent、GARP 3种模型对马铃薯甲虫在我国东北地区潜在分布风险进行了预测,并对预测结果进行了统计分析。GAM模型在本文设定数据条件下,表现出最好的预测正确性,其最优模型AUC达到0.87。由本研究过程可见,模型的选择首先应当在技术条件允许的情况下尽可能多的对模型在不同数据条件下表现出的性能进行考查,并对采用的分析结果保持足够的谨慎,在认识到模型可能存在局限的情况下有限度的接受和使用预测结果。
In this paper, taking potato beetle as an example, this paper explores the establishment of a quantitative assessment technique based on the predictions of different analytical methods and clarifies the suitable range and colonization risk of potato beetle in the northeast of China. Based on the unified environmental data and biological data, the potential risk of potato beetle in northeastern China was predicted by using GAM, Max Ent and GARP models respectively. The forecast results were statistically analyzed. The GAM model shows the best prediction accuracy under the data set in this paper, and its optimal model AUC reaches 0.87. It can be seen from the research process that the choice of model should first examine the performance of the model under different data conditions as much as possible under the technical conditions and be cautious about the analysis results adopted. After recognizing that the model There may be limited acceptance and use of forecast results.