On the importance of stationarity, strong baselines and benchmarks in transport prediction problems

Open in new window