Systematic Comparison of the Influence of Different Data Preprocessing Methods on the Performance of Gait Classifications Using Machine Learning

Autoren

Johannes Burdack
Fabian Horst
Sven Giesselbach
Ibrahim Hassan
Sabrina Daffner
Wolfgang I Schöllhorn

DOI

10.3389/fbioe.2020.00260

eISSN

2296-4185

Zeitschrift

Frontiers in Bioengineering and Biotechnology

Online publication date

2020

Status

Published online

Herausgeber

Frontiers Media SA

Herausgeber URL

http://dx.doi.org/10.3389/fbioe.2020.00260

Datum der Datenerfassung

2020

Titel

Systematic Comparison of the Influence of Different Data Preprocessing Methods on the Performance of Gait Classifications Using Machine Learning

Ausgabe der Zeitschrift

8

Datenquelle: Crossref

Abstract

Human movements are characterized by highly non-linear and multi-dimensional interactions within the motor system. Therefore, the future of human movement analysis requires procedures that enhance the classification of movement patterns into relevant groups and support practitioners in their decisions. In this regard, the use of data-driven techniques seems to be particularly suitable to generate classification models. Recently, an increasing emphasis on machine-learning applications has led to a significant contribution, e.g., in increasing the classification performance. In order to ensure the generalizability of the machine-learning models, different data preprocessing steps are usually carried out to process the measured raw data before the classifications. In the past, various methods have been used for each of these preprocessing steps. However, there are hardly any standard procedures or rather systematic comparisons of these different methods and their impact on the classification performance. Therefore, the aim of this analysis is to compare different combinations of commonly applied data preprocessing steps and test their effects on the classification performance of gait patterns. A publicly available dataset on intra-individual changes of gait patterns was used for this analysis. Forty-two healthy participants performed 6 sessions of 15 gait trials for 1 day. For each trial, two force plates recorded the three-dimensional ground reaction forces (GRFs). The data was preprocessed with the following steps: GRF filtering, time derivative, time normalization, data reduction, weight normalization and data scaling. Subsequently, combinations of all methods from each preprocessing step were analyzed by comparing their prediction performance in a six-session classification using Support Vector Machines, Random Forest Classifiers, Multi-Layer Perceptrons, and Convolutional Neural Networks. The results indicate that filtering GRF data and a supervised data reduction (e.g., using Principal Components Analysis) lead to increased prediction performance of the machine-learning classifiers. Interestingly, the weight normalization and the number of data points (above a certain minimum) in the time normalization does not have a substantial effect. In conclusion, the present results provide first domain-specific recommendations for commonly applied data preprocessing methods and might help to build more comparable and more robust classification models based on machine learning that are suitable for a practical application.

Addresses

Department of Training and Movement Science, Institute of Sport Science, Johannes Gutenberg-University, Mainz, Germany.

Autoren

Johannes Burdack
Fabian Horst
Sven Giesselbach
Ibrahim Hassan
Sabrina Daffner
Wolfgang I Schöllhorn

DOI

10.3389/fbioe.2020.00260

eISSN

2296-4185

Externe Identifier

PubMed Identifier: 32351945
PubMed Central ID: PMC7174559

Open access

true

ISSN

2296-4185

Zeitschrift

Frontiers in bioengineering and biotechnology

Sprache

eng

Medium

Electronic-eCollection

Online publication date

2020

Open access status

Open Access

Paginierung

260

Datum der Veröffentlichung

2020

Status

Published

Publisher licence

CC BY

Datum der Datenerfassung

2020

Titel

Systematic Comparison of the Influence of Different Data Preprocessing Methods on the Performance of Gait Classifications Using Machine Learning.

Sub types

research-article
Journal Article

Ausgabe der Zeitschrift

8

Files

https://www.frontiersin.org/articles/10.3389/fbioe.2020.00260/pdf https://europepmc.org/articles/PMC7174559?pdf=render

Datenquelle: Europe PubMed Central

Abstract

Human movements are characterized by highly non-linear and multi-dimensional interactions within the motor system. Therefore, the future of human movement analysis requires procedures that enhance the classification of movement patterns into relevant groups and support practitioners in their decisions. In this regard, the use of data-driven techniques seems to be particularly suitable to generate classification models. Recently, an increasing emphasis on machine-learning applications has led to a significant contribution, e.g., in increasing the classification performance. In order to ensure the generalizability of the machine-learning models, different data preprocessing steps are usually carried out to process the measured raw data before the classifications. In the past, various methods have been used for each of these preprocessing steps. However, there are hardly any standard procedures or rather systematic comparisons of these different methods and their impact on the classification performance. Therefore, the aim of this analysis is to compare different combinations of commonly applied data preprocessing steps and test their effects on the classification performance of gait patterns. A publicly available dataset on intra-individual changes of gait patterns was used for this analysis. Forty-two healthy participants performed 6 sessions of 15 gait trials for 1 day. For each trial, two force plates recorded the three-dimensional ground reaction forces (GRFs). The data was preprocessed with the following steps: GRF filtering, time derivative, time normalization, data reduction, weight normalization and data scaling. Subsequently, combinations of all methods from each preprocessing step were analyzed by comparing their prediction performance in a six-session classification using Support Vector Machines, Random Forest Classifiers, Multi-Layer Perceptrons, and Convolutional Neural Networks. The results indicate that filtering GRF data and a supervised data reduction (e.g., using Principal Components Analysis) lead to increased prediction performance of the machine-learning classifiers. Interestingly, the weight normalization and the number of data points (above a certain minimum) in the time normalization does not have a substantial effect. In conclusion, the present results provide first domain-specific recommendations for commonly applied data preprocessing methods and might help to build more comparable and more robust classification models based on machine learning that are suitable for a practical application.

Date of acceptance

2020

Autoren

Johannes Burdack
Fabian Horst
Sven Giesselbach
Ibrahim Hassan
Sabrina Daffner
Wolfgang I Schöllhorn

Autoren-URL

https://www.ncbi.nlm.nih.gov/pubmed/32351945

DOI

10.3389/fbioe.2020.00260

Externe Identifier

PubMed Central ID: PMC7174559

ISSN

2296-4185

Zeitschrift

Front Bioeng Biotechnol

Schlüsselwörter

convolutional neural network
data processing
data selection
gait classification
ground reaction force
multi-layer perceptron
random forest classifier
support vector machine

Sprache

eng

Country

Switzerland

Paginierung

260

Datum der Veröffentlichung

2020

Status

Published online

Titel

Systematic Comparison of the Influence of Different Data Preprocessing Methods on the Performance of Gait Classifications Using Machine Learning.

Sub types

Journal Article

Ausgabe der Zeitschrift

8

Datenquelle: PubMed

Author's licence

CC-BY

Autoren

Johannes Burdack
Fabian Horst
Sven Giesselbach
Ibrahim Hassan
Sabrina Daffner
Wolfgang I Schöllhorn

Hosting institution

Universitätsbibliothek Mainz

Sammlungen

JGU-Publikationen

Resource version

Published version

DOI

10.3389/fbioe.2020.00260

Funding acknowledgements

DFG, Open Access-Publizieren Universität Mainz / Universitätsmedizin Mainz

File(s) embargoed

false

Open access

true

ISSN

2296-4185

Zeitschrift

Frontiers in Bioengineering and Biotechnology

Schlüsselwörter

796 Sport
796 Athletic and outdoor sports and games

Sprache

eng

Open access status

Open Access

Paginierung

Art. 260

Datum der Veröffentlichung

2020

Public URL

https://openscience.ub.uni-mainz.de/handle/20.500.12030/5265

Herausgeber

Frontiers Media

Herausgeber URL

https://doi.org/10.3389/fbioe.2020.00260

Datum der Datenerfassung

2020

Datum, an dem der Datensatz öffentlich gemacht wurde

2020

Zugang

Public

Titel

Systematic comparison of the influence of different data preprocessing methods on the performance of gait classifications using machine learning

Ausgabe der Zeitschrift

8

Files

burdack_johannes-systematic_com-20201027094017196.pdf

Datenquelle: OPENSCIENCE.UB

Abstract

Human movements are characterized by highly non-linear and multi-dimensional interactions within the motor system. Recently, an increasing emphasis on machine-learning applications has led to a significant contribution to the field of gait analysis, e.g., in increasing the classification performance. In order to ensure the generalizability of the machine-learning models, different data preprocessing steps are usually carried out to process the measured raw data before the classifications. In the past, various methods have been used for each of these preprocessing steps. However, there are hardly any standard procedures or rather systematic comparisons of these different methods and their impact on the classification performance. Therefore, the aim of this analysis is to compare different combinations of commonly applied data preprocessing steps and test their effects on the classification performance of gait patterns. A publicly available dataset on intra-individual changes of gait patterns was used for this analysis. Forty-two healthy participants performed 6 sessions of 15 gait trials for 1 day. For each trial, two force plates recorded the 3D ground reaction forces (GRFs). The data was preprocessed with the following steps: GRF filtering, time derivative, time normalization, data reduction, weight normalization and data scaling. Subsequently, combinations of all methods from each preprocessing step were analyzed by comparing their prediction performance in a six-session classification using Support Vector Machines, Random Forest Classifiers, Multi-Layer Perceptrons, and Convolutional Neural Networks. In conclusion, the present results provide first domain-specific recommendations for commonly applied data preprocessing methods and might help to build more comparable and more robust classification models based on machine learning that are suitable for a practical application.

Autoren

Johannes Burdack
Fabian Horst
Sven Giesselbach
Ibrahim Hassan
Sabrina Daffner
Wolfgang I Schöllhorn

Autoren-URL

http://arxiv.org/abs/1911.04335v2

Zeitschrift

Front. Bioeng. Biotechnol.

Schlüsselwörter

cs.LG
cs.LG
stat.ML

Notes

12 pages, 3 figures, 4 tables

Paginierung

260

Datum der Veröffentlichung

2019

Herausgeber URL

http://dx.doi.org/10.3389/fbioe.2020.00260

Datum der Datenerfassung

2019

Datum, an dem der Datensatz öffentlich gemacht wurde

2019

Titel

Systematic Comparison of the Influence of Different Data Preprocessing Methods on the Performance of Gait Classifications Using Machine Learning

Ausgabe der Zeitschrift

8

Files

1911.04335v2.pdf

Datenquelle: arXiv

Abstract

Human movements are characterized by highly non-linear and multi-dimensional interactions within the motor system. Therefore, the future of human movement analysis requires procedures that enhance the classification of movement patterns into relevant groups and support practitioners in their decisions. In this regard, the use of data-driven techniques seems to be particularly suitable to generate classification models. Recently, an increasing emphasis on machine-learning applications has led to a significant contribution, e.g., in increasing the classification performance. In order to ensure the generalizability of the machine-learning models, different data preprocessing steps are usually carried out to process the measured raw data before the classifications. In the past, various methods have been used for each of these preprocessing steps. However, there are hardly any standard procedures or rather systematic comparisons of these different methods and their impact on the classification performance. Therefore, the aim of this analysis is to compare different combinations of commonly applied data preprocessing steps and test their effects on the classification performance of gait patterns. A publicly available dataset on intra-individual changes of gait patterns was used for this analysis. Forty-two healthy participants performed 6 sessions of 15 gait trials for 1 day. For each trial, two force plates recorded the three-dimensional ground reaction forces (GRFs). The data was preprocessed with the following steps: GRF filtering, time derivative, time normalization, data reduction, weight normalization and data scaling. Subsequently, combinations of all methods from each preprocessing step were analyzed by comparing their prediction performance in a six-session classification using Support Vector Machines, Random Forest Classifiers, Multi-Layer Perceptrons, and Convolutional Neural Networks. The results indicate that filtering GRF data and a supervised data reduction (e.g., using Principal Components Analysis) lead to increased prediction performance of the machine-learning classifiers. Interestingly, the weight normalization and the number of data points (above a certain minimum) in the time normalization does not have a substantial effect. In conclusion, the present results provide first domain-specific recommendations for commonly applied data preprocessing methods and might help to build more comparable and more robust classification models based on machine learning that are suitable for a practical application.

Autoren

Johannes Burdack
Fabian Horst
Sven Giesselbach
Ibrahim Hassan
Sabrina Daffner
Wolfgang I Schöllhorn

DOI

10.3389/fbioe.2020.00260

ISSN

2296-4185

Zeitschrift

Frontiers in bioengineering and biotechnology

Notes

file: http://www.ncbi.nlm.nih.gov/pubmed/32351945 file: http://www.ncbi.nlm.nih.gov/pubmed/32351945

Paginierung

260 - 260

Datum der Veröffentlichung

2020

Datum der Datenerfassung

2021

Titel

Systematic Comparison of the Influence of Different Data Preprocessing Methods on the Performance of Gait Classifications Using Machine Learning

Sub types

article

Ausgabe der Zeitschrift

8

Datenquelle: Manual

Systematic Comparison of the Influence of Different Data Preprocessing Methods on the Performance of Gait Classifications Using Machine Learning

Files

Files

Files

Werkzeuge