Publication

Abstract : Efficient Fuzzy Trajectory Clustering Algorithm tackles challenges in spatial trajectory data clustering by addressing varying sampling rates, serialization, uncertainty, redundancy, and spatial autocorrelation. Initially, it normalizes trajectories of varying lengths into equal lengths using the Douglas-Peucker algorithm. EFTCA then determines the optimal number of clusters using fuzzy partition coefficients. This fuzzy-based approach effectively handles uncertainty and redundancy in trajectory clustering. The algorithm’s performance is evaluated using standard cluster validity metrics, including the Adjusted Rand Index (ARI) and Fowlkes-Mallow Score (FMS), to compare EFTCA with conventional clustering techniques. The results reveal that EFTCA significantly outperforms standard methods, with ARI and FMS values approximately 1.31 times higher, demonstrating its superiority in handling complex trajectory data and providing more reliable and accurate clustering results. Proposed EFTCA is validated for accuracy against the researcher generated data as proof-of-concept validation.