Weighted dynamic time warping for optimal subsequence search
Name: Weighted dynamic time warping for optimal subsequence search
Journal: Tyumen 2019: 6th Conferenc
Authors: Alekseev Goncharov.
Abstract: The paper investigates the task of optimal subsequence search in high dimension time series. Author proposes to use a search with distance function that is stable in case of local and global time warpings: dynamic time warping method. It presents a baseline for solving the task. its weighted modification improves the quality and adds new properties: method lets to analyze the most significant measures. The experiment holds on the synthetic data based on real human activity data. The experiment results let us to use the method on the real seismological data for patterns search.