| Autor: |
Yazdizadeh, Ali, Patterson, Zachary, Farooq, Bilal |
| Zdroj: |
Journal of Big Data Analytics in Transportation; 20210101, Issue: Preprints p1-11, 11p |
| Abstrakt: |
This study experiments with the use of adversarial networks to classify travel mode based on one-dimensional smartphone trajectory data. We use data from a large-scale smartphone travel survey in Montreal, Canada. We convert GPS trajectories into fixed-sized segments with five channels (or variables). We develop different GANs architectures and compare their prediction results with convolutional neural networks (CNNs). The best semi-supervised GANs model led to a prediction accuracy of 83.4%, while the best CNN model was able to achieve the prediction accuracy of 81.3%. The results compare favorably with previous studies, especially when taking the large, real-world nature of the dataset into account. The developed semi-supervised GANs models share the same architectural innovations used in the image recognition literature, that we show can be used in travel information inference from smartphone travel survey data, not only to generate more labeled samples but also to improve the prediction performance of the classifier. Future work will allow exploration of better-performing models either with more channels and/or improved architectures. |
| Databáze: |
Supplemental Index |
| Externí odkaz: |
|
Full text from SpringerLink