The hot spots of people's activities are an important part of urban spatial structure research, and the analysis of the driving forces of urban hot spots can provide a scientific basis for urban planning. The determination of urban hot spots is affected by the heterogeneity of the analysis data, resulting in differences in the hot spot driving factors. To analyze the spatial distributions of urban hot spots, this article conducted a series of exploratory spatial analyses on spatiotemporal travel data and further applied a binary logistic regression (BLR) model to analyze the driving factors affecting the distribution of urban hot spots to accurately characterize the distribution pattern of urban hot spots. First, a spatial autocorrelation test was performed on the trajectory data. Based on the test results, the Getis–Ord statistical method was used to calculate the distribution area of urban hot spots. Second, considering the internal and external factors that affect the distribution of urban hot spots, 10 driving factors in the categories of terrain, socioeconomics, and road and transportation facility accessibility were selected as the independent variables in the regression model of the hot spot distribution. Third, a BLR model was used to establish a driving force model that altered the distribution of hot spots in cities to explore the influencing factors and the degree of the hot spot distribution. Finally, the accuracy of the hot spot driving force model was evaluated according to the receiver operating characteristic (ROC) curve. Taking Wuhan as an example, the results show that the area under the ROC curve of the hot spot model established in this article reaches 0.918, and the model has a good fit. The formation of urban hot spots is a combination of internal driving factors such as terrain and socioeconomics as well as external driving factors such as the accessibility of roads and transportation facilities. Among these factors, transportation facility accessibility contributes most to the hot spot distribution.

中文翻译：

---

基于探索性分析和二元逻辑回归模型的城市热点驱动力研究

人们活动热点是城市空间结构研究的重要组成部分，对城市热点驱动力的分析可以为城市规划提供科学依据。城市热点的确定受分析数据的异质性影响，导致热点驱动因素存在差异。为分析城市热点的空间分布，本文对时空出行数据进行了一系列探索性空间分析，并进一步应用二元逻辑回归（BLR）模型分析影响城市热点分布的驱动因素，以准确刻画城市热点分布特征。城市热点分布格局。首先，对轨迹数据进行空间自相关测试。根据测试结果，Getis-Ord采用统计方法计算城市热点分布面积。其次，综合考虑影响城市热点分布的内外部因素，选取地形、社会经济、道路和交通设施可达性等10个驱动因素作为热点分布回归模型的自变量。第三，利用BLR模型建立改变城市热点分布的驱动力模型，探索热点分布的影响因素和程度。最后，根据受试者工作特征（ROC）曲线评估热点驱动力模型的准确性。以武汉为例，结果表明，本文建立的热点模型的ROC曲线下面积达到0.918，模型拟合良好。城市热点的形成是地形、社会经济等内在驱动因素与道路、交通设施可达性等外在驱动因素的综合作用。在这些因素中，交通设施可达性对热点分布的贡献最大。