The AI team of the DeustoTech Mobility Unit ranked 3rd in the TRANSFOR19 competition
Published on Thursday, 24 January 2019 - 07:44
The Artificial Intelligence research team of the DeustoTech’s Mobility Unit ranked 3rd in the Transportation Forecasting Competition, TRANSFOR 19. This challenge was organized by the Standing Committee on Artificial Intelligence and Advanced Computing Applications (ABJ70) of the Transportation Research Board, supported by the IEEE ITSS Technical Activities Sub-Committee “Smart Cities and Smart Mobility” and sponsored by Didi Chuxing.
The objective of the competition was to evaluate the performance of statistical and computational intelligence methods in short-term traffic forecasting. The competition was run in two stages. In the first stage, the team from Deusto, composed of Toon Bogaerts, Antonio D. Masegosa, Juan S. Angarita, Pedro Lopez-Garcia, Amgad Elsayed and Enrique Onieva, was selected among the top five finalists.
The second stage of the competition was held in Washington on January 13th, 2019 at the Transportation Research Board (TRB) 98th Annual Meeting. The 5 finalists presented their work during the TRB Annual Meeting Workshop 1058 on “Big Data Without Machine Learning Is Just Lots of Data: A Guided Tour to Big Data and Machine Learning” organized by ABJ70.
Toon Bogaerts presented DestoTech’s Mobility's work in Washington
The results of the competition can be consulted here. The Deusto team,represented by Toon Bogaerts, ranked in third position only surpassed by theTraffic Flaw Theory Team (University of California, USA) and IBM Team (Ireland). For the final decision, the member of the panels took into account four criteria: i. Accuracy (50%), ii. Degree on novelty (25%), iii. Quality of the code (10%), and iv. Quality of the presentation (15%).Taking into account the international prestige of the institutions of the other participating teams, the third position achieved by the Deusto’s team is very meritorious. The interested reader can have more details about the competition and the submissions from all team here.