Trip destination prediction is an area of increasing importance in many applications such as trip planning, autonomous driving and electric vehicles . In this paper, we present a unified framework for trip destination prediction in an online setting . We investigate the different configurations of clustering algorithms and prediction models on a real-world dataset . By using traditional clusteringmetrics and accuracy, we demonstrate that both the clustering and the entireframework yield consistent results compared to the offline setting. Finally, we propose a novel regret metric for evaluating the entire online framework incomparison to its offline counterpart. This metric makes it possible to relatethe source of erroneous predictions to either the clustered or the predictionmodel . This metric, we show that the proposed methods converge to aprobability distribution resembling the true underlying distribution and enjoy a lower regret than all of the baselines. Using this metric, We show that proposed methods converged to a ProbabilityDistribution of

Author(s) : Victor Eberstein, Jonas Sjöblom, Nikolce Murgovski, Morteza Haghir Chehreghani

Links : PDF - Abstract

Code :

https://github.com/doty-k/world_models


Coursera

Keywords : trip - prediction - metric - framework - destination -

Leave a Reply

Your email address will not be published. Required fields are marked *