Activity imputation for trip-chains elicited from smart-card data using a continuous hidden Markov model
详细信息 查看全文
- 作者：Gain Han a9yness@gmail.com" class="auth_mail" title="E-mail the corresponding author ; Keemin Sohn ; kmsohn@cau.ac.kr" class="auth_mail" title="E-mail the corresponding author
- 关键词：Smart-card data ; Activity-based model ; Activity imputation ; Continuous hidden Markov model ; Machine learning ; Trip chain
- 刊名：Transportation Research Part B: Methodological
- 出版年：2016
- 出版时间：January 2016
- 年：2016
- 卷：83
- 期：Complete
- 页码：121-135
- 全文大小：393 K
文摘
- •
The present study proposed a rigorous methodology to impute the sequence of activities elicited form smart-card data using a continuous hidden Markov model (CHMM).
- •
The proposed model requires neither labeled data for training nor subsequent measurements such as prompted-recall surveys.
- •
The present study showed the power of unsupervised machine-learning models.
- •
Self-clustered activities and transition probabilities between them were fully validated by observed data.