Comparison of Advanced GPS Data Imputation Algorithms for Detection of Transportation Mode and Activity Episode
Comparison of Advanced GPS Data Imputation Algorithms for Detection of Transportation Mode and Activity Episode
复制标题
DOI：
--
复制
发表时间：
2014
期刊：
影响因子：
--
通讯作者：
T. Feng;H. Timmermans
中科院分区：
文献类型：
--
作者：
T. Feng;H. Timmermans
GPS（全球定位系统）数据收集越来越多地被认为是传统旅行调查方法收集活动旅行数据的替代方法。从非正式的临时决策规则到先进的机器学习方法，各种算法已被应用于从 GPS 轨迹中提取活动-旅行模式。然而，这些不同算法的准确性很难比较，因为它取决于使用 GPS 的空间环境、识别的交通模式的数量、输入变量的类型以及用于验证的数据。因此，本文的目的是系统地比较用于检测交通方式和活动事件的不同算法的相对性能。具体来说，根据总体错误率和命中率，对相同数据选择朴素贝叶斯、贝叶斯网络、逻辑回归、多层感知器、支持向量机、决策表和C4.5算法进行比较。结果表明，在训练数据和测试数据的正确识别实例的百分比和 Kappa 值方面，贝叶斯网络比其他算法具有更好的性能，因为贝叶斯网络在 GPS 数据插补的背景下相对有效且可推广。
GPS (Global Positioning System) data collection has been increasingly considered as an alternative to traditional travel survey methods to collect activity-travel data. Algorithms varying from the informal ad-hoc decision rules to advanced machine learning methods have been applied to extract activity-travel patterns from GPS traces. However, the accuracy of these different algorithms is difficult to compare as it depends on the spatial context in which GPS is used, the number of identified transportation modes, type of input variables and data used for validation. The aim of this paper therefore is to systematically compare the relative performance of different algorithms for the detection of transportation modes and activity episode. In particular, the naive Bayesian, Bayesian network, logistic regression, multilayer perceptron, support vector machine, decision table and C4.5 algorithms are selected and compared for the same data according to the overall error rates and hit ratios. Results show that the Bayesian network has a better performance than the other algorithms in terms of the percentage of correctly identified instances and Kappa values for both the training data and test data in the sense that Bayesian network is relatively efficient and generalizable in the context of GPS data imputation.