A Survey and Comparison of Trajectory Classification Methods

A Deep Learning Streaming Methodology for Trajectory Classification

“…On the other hand, trajectory clustering approaches are often employed to form groups of AIS positions with similar spatiotemporal behaviors, uncovering behaviors that are harder to predefine. Although there is an abundance of studies in the literature regarding offline trajectory classification and clustering [1][2][3][4][5], fewer works have focused on steam processing of events in the maritime domain [6][7][8][9][10]. Event processing methodologies are faced with significant challenges when employed on streaming data where the requirements for such applications demand low memory consumption and decreased latencies.…”

“…marinetraffic.com/hc/en-us/articles/217631867-How-often-do-the-positions-of-thevessels-get-updated-on-MarineTraffic-, accessed on 7 April 2021). This is in contrast to time-series methodologies [2] that are inherently unsuitable for such tasks and require data points at fixed time points; • Trajectory classification approaches found in the literature [4,10,14,15] require a preprocessing step such as the understanding and analysis of data and the selection of features suitable only for the mobility patterns to be classified. This means that features selected for a certain mobility pattern cannot be applied to other patterns as well [14].…”

Evaluating the effect of compressing algorithms for trajectory similarity and classification problems

“…In the classification tasks, after the extraction of the features, the Random Forest (RF) is trained and used to compare the techniques, once it is fast and commonly used by the techniques in the literature. The works of [34] and [47] have compared several classifiers including SVM (Support Vector Machine), MLP (Multi Layer Perceptron) and random forest, and they have shown a similar accuracy. We have selected RF because it performs faster Dodge [31], Zheng [28] and Xiao [32] Space and time Numerical features, e.g.…”

“…The main difference between the trajectory classification techniques is the type of trajectory features they extract for training the classification model [24]. Most works in trajectory classification extract features from the spatio-temporal properties of trajectories, as the speed, acceleration, direction change, etc.…”

“…In the following we present a detailed description of the trajectory datasets. Many of these datasets have already being used in [24] and [34] for comparing several classification methods.…”