Improving alighting stop inference accuracy in the trip chaining method using neural networks · Study