航班动态定价机制下的机票价格序列变点估计

文章旨在研究受航空业动态定价机制影响下的机票价格序列变点估计模型，文中分析了机票价格u8序列数据的结构特点，提出了可用于高噪声数据环境下、阶梯状、带明显多变点的多阶段序列变点估计框架，该框架中级联组合了DBSCAN算法、EM-高斯混合模型聚类、凝聚层次聚类算法和基于乘积划分模型的变点估计方法等多种成熟的数据分析方法，通过对“北京-昆明”航线航班的实证分析，验证了数据分析框架的有效性和普遍适用性。

Change Point Estimation for airfare Sequence under Dynamic Pricing Strategy in Air Industry

This paper aims to present change point estimation model for airfare sequence under dynamic pricing strategy in Air Industry. This paper analyses structural characteristics of ticket price series and present a change point estimation framework which can be applied to step-like multi-stage series with high noise data and multiple change points. This framework uses a combination of DBSCAN algorithm, EM-Gaussian Mixture Model clustering, hierarchical clustering algorithm and Product Partition Model for change point detection. By empirical studies on the flights for “Beijing to Kunming” route, this paper verifies the validity and general applicability of the data analysis framework.