E-book
Requirements for data sources and their quality for the implementation of a system for predicting traffic conditions based on machine learning
Requirements for data sources and their quality for the implementation of a system for predicting traffic conditions based on machine learning
Couldn't load pickup availability
The methodology refers to activities and actors related to the preparation of data for the possibility of using a neural network to predict the development of the traffic flow. It is therefore intended for road administrators who have data and information and will be interested in using these tools, as well as for private entities who have road traffic data and are interested in preparing data for use in a deep neural network model.
Detail
Author
- Rozhon J.
- Mikolašek I.
- Mynarik J.
- Uherka Z.
- Švédová Z.
- Bambušek M.
- Safařík J.
- Novobilsky J.
- Ščerba M.
Content
1. General description of the machine learning method
2. Data collection
3. Data preparation for machine learning
4. Basic content-structural requirements for data availability
5. Data format and structure
6. Requirements for documentation of data sources
7. Review and Validation of Data
8. Factors entering traffic flow models based on machine learning
9. Comparison of the novelty of procedures
10. Economic aspects
Dedication
This methodology was co-financed with the state support of the Technology Agency of the Czech Republic and the Ministry of Transport within the Transport 2020+ program, as part of the project solution CK01000139 System for predicting the development of traffic flow dynamics based on a deep neural network.
www.tacr.cz
Colophon
ISBN 978-80-88655-08-4