Local Proportional Scaling of Time-Series Data: Method and Applications

in Motor Control
Restricted access

Purchase article

USD $24.95

Student 1 year subscription

USD $76.00

1 year subscription

USD $101.00

Student 2 year subscription

USD $144.00

2 year subscription

USD $188.00

A method for analysis of time-series data, local proportional scaling (LPS), is proposed and its applications in motor control and biomechanics are discussed. The method is based on comparison of two time curves: a reference curve x(t) and a test curve x'(t'). By assumption, x'(t') is received from x(t) by local affine transformations, local extensions/compressions along the x and t axes [x(t)→x'(t'), where → stands for the local extensions/compressions along the x and t axes]. The aim of the LPS method is to discover the underlying transformations, including gain indexes, time epochs, velocity quotients, time segments, and time quotients. The LPS method can be used for (a) comparing the time-series curves in a concise transparent manner; (b) scaling the curves, bringing x'(t') in conformity with x(t); (c) automatic segmentation of the time series data; and (d) data classification.

K. Kanatani-Fujimoto and Vladimir M. Zatsiorsky are with the Biomechanics Laboratory, Department of Kinesiology, The Pennsylvania State University, University Park, PA 16802. B.V. Lazareva is with the Biostatistics Unit, Cornell University, Ithaca, NY 14853. Direct correspondence to V.M. Zatsiorsky.

Motor Control
Article Metrics
All Time Past Year Past 30 Days
Abstract Views 12 12 2
Full Text Views 1 1 0
PDF Downloads 2 2 0
Altmetric Badge
PubMed
Google Scholar