Physical Activity Recognition from Accelerometer Data Using a Multi-Scale Ensemble Method

Accurate and detailed measurement of an individual’s physical activity is a key requirement for helping researchers understand the relationship between physical activity and health. Accelerometers have become the method of choice for measuring physical activity due to their small size, low cost, convenience and their ability to provide objective information about physical activity. However, interpreting accelerometer data once it has been collected can be challenging. In this work, we applied machine learning algorithms to the task of physical activity recognition from triaxial accelerometer data. We employed a simple but effective approach of dividing the accelerometer data into short non-overlapping windows, converting each window into a feature vector, and treating each feature vector as an i.i.d training instance for a supervised learning algorithm. In addition, we improved on this simple approach with a multi-scale ensemble method that did not need to commit to a single window size and was able to leverage the fact that physical activities produced time series with repetitive patterns and discriminative features for physical activity occurred at different temporal scales.

• Publication year

  2013

• Venue

  Conference on Innovative Applications of Artificial Intelligence

• Publication date

  2013-07-14

• Fields of study

  Medicine, Computer Science, Engineering

• Identifiers
  DOI 10.1609/aaai.v27i2.18997
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Motion primitive-based human activity recognition using a bag-of-features approach

  2012cited by this paper
• Calibration and validation of wearable monitors.

  2012cited by this paper
• Artificial neural networks to predict activity type and energy expenditure in youth.

  2012influential reference
• HASC Challenge: gathering large scale human activity corpus for the real-world activity understandings

  2011cited by this paper
• LIBSVM: A library for support vector machines

  2011cited by this paper
• Evaluation of neural networks to identify types of activity using accelerometers.

  2011cited by this paper
• Modern Applied Statistics with S

  2010cited by this paper
• A brief survey on sequence classification

  2010cited by this paper
• Experimental comparison of representation methods and distance measures for time series data

  2010influential reference
• Detection of type, duration, and intensity of physical activity using an accelerometer.

  2009cited by this paper
• An artificial neural network to estimate physical activity energy expenditure and identify physical activity type from an accelerometer.

  2009cited by this paper
• Time series shapelets: a new primitive for data mining

  2009cited by this paper
• Physical activity measurement- a primer for health promotion

  2006cited by this paper
• A Hybrid Discriminative/Generative Approach for Modeling Human Activities

  2005cited by this paper
• Estimation of Walking Energy Expenditure by Using Support Vector Regression

  2005cited by this paper
• Conducting accelerometer-based activity assessments in field-based research.

  2005cited by this paper
• Activity Recognition from Accelerometer Data

  2005cited by this paper
• Activity Recognition from User-Annotated Acceleration Data

  2004cited by this paper
• Defining accelerometer thresholds for activity intensities in adolescent girls.

  2004cited by this paper
• A symbolic representation of time series, with implications for streaming algorithms

  2003cited by this paper
• On the Need for Time Series Data Mining Benchmarks: A Survey and Empirical Demonstration

  2002cited by this paper
• Challenges and Opportunities for Measuring Physical Activity in Sedentary Adults

  2001cited by this paper
• Stacked generalization

  1992cited by this paper
• Medicine and Science in Sports and Exercise

  1981cited by this paper

• Machine Learning to Identify a New Digital Biomarker to Monitor Everyday Upper Limb Use in Children with Unilateral Cerebral Palsy

  2026influential citation
• Advancing standardized cattle behavior classification with a random forest model evaluated across diverse datasets

  2026cites this paper
• Enhancing Sensor-Based Human Physical Activity Recognition Using Deep Neural Networks

  2025cites this paper
• A smartphone-based human activities recognition using novel multi-stream movelets on fusion of accelerometer and gyroscope data and classification using different distance metrics

  2024cites this paper
• Adherence to the Mediterranean Diet in Children and Adolescents and Association with Multiple Outcomes: An Umbrella Review

  2024cites this paper
• A stacked CNN and random forest ensemble architecture for complex nursing activity recognition and nurse identification

  2024cites this paper
• Cluster Analysis to Find Temporal Physical Activity Patterns Among US Adults

  2023cites this paper
• Overview of Human Activity Recognition Using Sensor Data

  2023cites this paper
• A Tutorial on Internet of Behaviors: Concept, Architecture, Technology, Applications, and Challenges

  2023cites this paper
• Sensor-Based Human Activity and Behavior Computing

  2021cites this paper
• Internet of Things Enabled Smart Animal Farm Prototype

  2021cites this paper
• The Determinants of Health-Related Quality of Life in a Sample of Primary School Children: A Cross-Sectional Analysis

  2021cites this paper
• Motion-Based Activities Monitoring through Biometric Sensors Using Genetic Algorithm

  2021cites this paper
• Towards human activity recognition for ubiquitous health care using data from awaist-mounted smartphone

  2020cites this paper
• A comparison of different machine learning algorithms, types and placements of activity monitors for physical activity classification

  2020cites this paper
• Survey on wearable sensor modality centred human activity recognition in health care

  2020cites this paper
• Automated Detection of Exercise Sessions in Patients With Peripheral Artery Disease: EVIDENCE FOR AN EXERCISE DOSE RESPONSE TO TRAINING.

  2020cites this paper
• A Novel Accelerometer Mounting Method for Sensing Performance Improvement in Acoustic Measurements From the Knee.

  2020cites this paper
• Associations between in-school-hours physical activity and child health-related quality of life: A cross-sectional study in a sample of Australian primary school children

  2020cites this paper
• PARCIV: Recognizing physical activities having complex interclass variations using semantic data of smartphone

  2020cites this paper
• A Comprehensive Study of Smartphone-Based Indoor Activity Recognition via Xgboost

  2019cites this paper
• Personalized Sleep Parameters Estimation from Actigraphy: A Machine Learning Approach

  2019cites this paper
• Human Activity Recognition Based on Motion Sensor Using U-Net

  2019cites this paper
• Registration and Analysis of Acceleration Data to Recognize Physical Activity

  2019cites this paper
• NSS_A_220716 387..399

  2019cites this paper
• Predicting Daily Activities Effectiveness Using Base-level and Meta level Classifiers

  2019cites this paper
• Recognizing Multi-Agent Plans When Action Models and Team Plans Are Both Incomplete

  2019cites this paper
• A quantitative comparison of Overlapping and Non-overlapping sliding windows effects for human activity recognition using inertial sensors

  2019cites this paper
• Multi-Sensor Activity Monitoring: Combination of Models with Class-Specific Voting

  2019cites this paper
• A survey on wearable sensor modality centred human activity recognition in health care

  2019cites this paper
• The purpose driven privacy preservation for accelerometer-based activity recognition

  2018cites this paper
• Investigation on the Effect of Different Window Size in Segmentation for Common Sport Activity

  2018cites this paper
• Smart home for elder care using wireless sensor

  2018cites this paper
• Activity Recognition with Evolving Data Streams

  2018cites this paper
• Discovering Underlying Plans Based on Shallow Models

  2018cites this paper
• Development and application of a machine learning algorithm for classification of elasmobranch behaviour from accelerometry data

  2018cites this paper
• Toward personalized sleep-wake prediction from actigraphy

  2018cites this paper
• Accurate and efficient human activity recognition

  2018cites this paper
• Attributes’ Importance for Zero-Shot Pose-Classification Based on Wearable Sensors

  2018cites this paper
• Human activity recognition based on time series analysis using U-Net

  2018cites this paper
• Recommender Systems for Healthy Behavior Change

  2017cites this paper
• CNN based approach for activity recognition using a wrist-worn accelerometer

  2017cites this paper
• Learning Activity Predictors from Sensor Data: Algorithms, Evaluation, and Applications

  2017cites this paper
• Activity Learning as a Foundation for Security Monitoring in Smart Homes

  2017cites this paper
• Sleep/wake state prediction and sleep parameter estimation using unsupervised classification via clustering

  2017cites this paper
• Thyme: Improving Smartphone Prompt Timing Through Activity Awareness

  2017cites this paper
• kNN Sampling for Personalised Human Activity Recognition

  2017cites this paper
• Activity and health status monitoring system

  2017cites this paper
• Augmenting Wearable Sensor Data with Physical Constraint for DNN-Based Human-Action Recognition

  2017cites this paper
• Learning Deep Features for kNN-Based Human Activity Recognition

  2017cites this paper
• Forecasting occurrences of activities

  2017cites this paper
• StraightenUp: Implementation and Evaluation of a Spine Posture Wearable

  2017cites this paper
• A smart segmentation technique towards improved infrequent non-speech gestural activity recognition model

  2017cites this paper
• IoT at the grassroots — Exploring the use of sensors for livestock monitoring

  2017influential citation
• Markov Dynamic Subsequence Ensemble for Energy-Efficient Activity Recognition

  2017influential citation
• IoT at the grassroots — Exploring the use of sensors for livestock monitoring

  2017cites this paper
• Feature Selection for Activity Recognition from Smartphone Accelerometer Data

  2017cites this paper
• Title: A decision support system for self-management of low back pain Activity: PHC-28-2015 Predictive modelling RIA

  2017cites this paper
• Detection of the smart phone position on user using inertial sensors

  2016cites this paper
• Motion based remote camera control with mobile devices

  2016cites this paper
• Comparative evaluation of features and techniques for identifying activity type and estimating energy cost from accelerometer data

  2016cites this paper
• Multicolumn Bidirectional Long Short-Term Memory for Mobile Devices-Based Human Activity Recognition

  2016cites this paper
• Exploring the Application of Hybrid Evolutionary Computation Techniques to Physical Activity Recognition

  2016cites this paper
• Towards unsupervised physical activity recognition using smartphone accelerometers

  2016cites this paper
• Recognizing Complex Activities by a Probabilistic Interval-Based Model

  2016cites this paper
• Utilisation des Random Forests pour la reconnaissance d'activité humaine: une étude complète avec le jeu de données d'Opportunity

  2016cites this paper
• Detecting Human Movement Patterns Through Data Provided by Accelerometers. A Case Study Regarding Alzheimer's Disease

  2016cites this paper
• Efficient Multi-Instance Learning for Activity Recognition from Time Series Data Using an Auto-Regressive Hidden Markov Model

  2016cites this paper
• A Comparative Study on Human Activity Recognition Using Inertial Sensors in a Smartphone

  2016cites this paper
• Using Random Forest for Opportunistic Human Activity Recognition: a complete study on Opportunity dataset

  2016cites this paper
• Talking Hands: RNN-based Sign Language Recognition

  2016influential citation
• SELFBACK - Activity Recognition for Self-management of Low Back Pain

  2016cites this paper
• A review of sensor selection, sensor devices and sensor deployment for wearable sensor-based human activity recognition systems

  2016cites this paper
• Motion Primitive Forests for Human Activity Recognition Using Wearable Sensors

  2016cites this paper
• A novel method for short-time human activity recognition based on improved template matching technique

  2016cites this paper
• Prediction of inhabitant activities in smart environments

  2015cites this paper
• Subject-dependent physical activity recognition using single sensor accelerometer

  2015cites this paper
• Wearable sensors network for activity recognition using inertial sensors

  2015cites this paper
• Toward Learning and Mining from Uncertain Time-Series Data for Activity Prediction

  2015cites this paper
• Hierarchical algorithm for daily activity recognition via smartphone sensors

  2015cites this paper
• Efficient Parameterized Methods for Physical Activity Detection using only Smartphone Sensors

  2015influential citation
• Feature Set Optimization for Physical Activity Recognition Using Genetic Algorithms

  2015cites this paper
• Physical Activity Classification with Conditional Random Fields

  2015cites this paper
• Data-Driven Activity Prediction: Algorithms, Evaluation Methodology, and Applications

  2015cites this paper
• The Elderly’s Independent Living in Smart Homes: A Characterization of Activities and Sensing Infrastructure Survey to Facilitate Services Development

  2015cites this paper
• Predicting human movement with multiple accelerometers using movelets.

  2014cites this paper
• Learning to detect subway arrivals for passengers on a train

  2014cites this paper
• Dynamic Programming for Instance Annotation in Multi-Instance Multi-Label Learning

  2014cites this paper
• Proceedings of the Twenty-Eighth AAAI Conference on Artificial Intelligence

  2014cites this paper
• Human Activity Recognition from Triaxial Accelerometer Data - Feature Extraction and Selection Methods for Clustering of Physical Activities

  2014cites this paper
• A Time Series Classification Approach for Motion Analysis Using Ensembles in Ubiquitous Healthcare Systems

  2014cites this paper
• Window Size Impact in Human Activity Recognition

  2014cites this paper
• Decomposing Activities of Daily Living to Discover Routine Clusters

  2014cites this paper
• Multivariate linear regression based activity recognition and classification

  2014cites this paper
• Human Activity Data Discovery based on Accelerometry

  2013cites this paper