We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.LG

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Machine Learning

Title: A novel activity pattern generation incorporating deep learning for transport demand models

Abstract: Activity generation plays an important role in activity-based demand modelling systems. While machine learning, especially deep learning, has been increasingly used for mode choice and traffic flow prediction, much less research exploiting the advantage of deep learning for activity generation tasks. This paper proposes a novel activity pattern generation framework by incorporating deep learning with travel domain knowledge. We model each activity schedule as one primary activity tour and several secondary activity tours. We then develop different deep neural networks with entity embedding and random forest models to classify activity type, as well as to predict activity times. The proposed framework can capture the activity patterns for individuals in both training and validation sets. Results show high accuracy for the start time and end time of work and school activities. The framework also replicates the start time patterns of stop-before and stop-after primary work activity well. This provides a promising direction to deploy advanced machine learning methods to generate more reliable activity-travel patterns for transport demand systems and their applications.
Comments: 21 pages, 12 figures
Subjects: Machine Learning (cs.LG); Multiagent Systems (cs.MA)
Cite as: arXiv:2104.02278 [cs.LG]
  (or arXiv:2104.02278v1 [cs.LG] for this version)

Submission history

From: The Danh Phan [view email]
[v1] Tue, 6 Apr 2021 04:07:05 GMT (2348kb,D)

Link back to: arXiv, form interface, contact.