Formed trajectories of sets of points.Experimented on finding similarities between trajectories based on DTW (Dynamic Time Warping) and LCSS (Longest Common SubSequence) algorithms.Modeled trajectories as strings based on a Grid representation.Benchmarked KNN, Random Forest, Logistic Regression classification algorithms to classify efficiently trajectories.
A university project for the postgraduate class of Data Mining.
We were given a train_set with geographical points paired with the time interval. Firstly, we cleaned the dataset and then we formed t…