Published on Mon Dec 21 2015

Multivariate Time Series Classification Using Dynamic Time Warping Template Selection for Human Activity Recognition

Skyler Seto, Wenyu Zhang, Yichen Zhou

We propose a template selection approach based on Dynamic Time Warping. We demonstrate the predictive capability of the algorithm on both simulated and real smartphone data.

0
0
0
Abstract

Accurate and computationally efficient means for classifying human activities have been the subject of extensive research efforts. Most current research focuses on extracting complex features to achieve high classification accuracy. We propose a template selection approach based on Dynamic Time Warping, such that complex feature extraction and domain knowledge is avoided. We demonstrate the predictive capability of the algorithm on both simulated and real smartphone data.

Thu Apr 22 2021
Machine Learning
A Feature Selection Method for Multi-Dimension Time-Series Data
0
0
0
Thu Aug 24 2017
Computer Vision
Gait Recognition from Motion Capture Data
Gait recognition from motion capture data can be improved by the use of machine learning. This research makes motion capture technology directly applicable for human recognition. Results indicate a highPortability of learned features.
0
0
0
Wed Jan 15 2014
Artificial Intelligence
Highly comparative feature-based time-series classification
A highly comparative, feature-based approach to time series classification has been introduced. It uses an extensive database of algorithms to extract thousands of interpretable features from time series. The resulting classifiers automatically learn the differences between classes using a reduced number of time series properties.
0
0
0
Thu Nov 30 2017
Machine Learning
Multivariate Time Series Classification with WEASEL+MUSE
WEASEL+MUSE builds a multivariate feature vector using a sliding-window approach. It then extracts discrete features per window and dimension. The feature vector is subsequently fed through feature selection, removing non-discriminative features, and analysed by a machine learning classifier.
0
0
0
Mon Mar 04 2019
Machine Learning
Making the Dynamic Time Warping Distance Warping-Invariant
The literature postulates that the dynamic time warping (dtw) distance can't cope with temporal variations. We convert the dtw-distance to a warping-invariant semi-metric. The twi-distance requires less storage and computation time.
0
0
0
Mon Dec 18 2017
Machine Learning
A Shapelet Transform for Multivariate Time Series Classification
Shapelets are phase independent subsequences designed for time series classification. We propose three adaptations to the Shapelet Transform (ST) to capture multivariate features. We demonstrate that multivariate shapelets are not significantly worse than other state-of-the-art algorithms.
0
0
0