https://www.kaggle.com/c/pkdd-15-predict-taxi-service-trajectory-i
https://www.kaggle.com/c/pkdd-15-predict-taxi-service-trajectory-i/data
Experimented in: Word2vec, K-means, Mean Shift clustering, Deep Learning
- Using k-means algorithm for clustering destination points.
- Using the word2vec model to embed each point in the map, datetime, and taxiId to a real numeric vector.
- Using multi-layer perceptron (Deep Learning) to predict the destination of the taxi trip to one of clusters.
- Using reduce_mean of points in the cluster to predict final destination of the taxi trip.