Bivariate change point detection in movement direction and speed

• DOI: 10.48550/arxiv.2402.02489
• Publication Date: 2024-02-04
• Authors: Solveig Plomer; Theresa Ernst; Philipp Gebhardt; Enrico Schleiff; Ralph Neininger; Gaby Schneider
• Keywords: Methodology (stat.ME); FOS: Computer and information sciences; FOS: Biological sciences; Primary 62P10, 60F17, Secondary: 60G15, 60G50; FOS: Mathematics; Mathematics - Statistics Theory; Applications (stat.AP); Statistics Theory (math.ST); Quantitative Biology - Quantitative Methods; Statistics - Applications; Statistics - Methodology; Quantitative Methods (q-bio.QM)

Biological movement patterns can sometimes be quasi linear with abrupt changes in direction and speed, as plastids root cells investigated here. For the analysis of such we propose a new stochastic model for along structures. Maximum likelihood estimators are provided, due to serial dependencies increments, classical MOSUM statistic is replaced by moving kernel estimator. Convergence resulting difference process strong consistency variance estimator shown. We estimate change points graphical technique distinguish between speed.