Hierarchical classification method based on selective learning of slacked hierarchy for activity recognition systems

Abstract Physical activity recognition using wearable sensors has gained significant interest from researchers working in the field of ambient intelligence and human behavior analysis. The problem of multi-class classification is an important issue in the applications which naturally has more than two classes. A well-known strategy to convert a multi-class classification problem into binary sub-problems is the error-correcting output coding (ECOC) method. Since existing methods use a single classifier with ECOC without considering the dependency among multiple classifiers, it often fails to generalize the performance and parameters in a real-life application, where different numbers of devices, sensors and sampling rates are used. To address this problem, we propose a unique hierarchical classification model based on the combination of two base binary classifiers using selective learning of slacked hierarchy and integrating the training of binary classifiers into a unified objective function. Our method maps the multi-class classification problem to multi-level classification. A multi-tier voting scheme has been introduced to provide a final classification label at each level of the solicited model. The proposed method is evaluated on two publicly available datasets and compared with independent base classifiers. Furthermore, it has also been tested on real-life sensor readings for 3 different subjects to recognize four activities i.e. Walking, Standing, Jogging and Sitting. The presented method uses same hierarchical levels and parameters to achieve better performance on all three datasets having different number of devices, sensors and sampling rates. The average accuracies on publicly available dataset and real-life sensor readings were recorded to be 95% and 85%, respectively. The experimental results validate the effectiveness and generality of the proposed method in terms of performance and parameters.

[1]  Koby Crammer,et al.  On the Learnability and Design of Output Codes for Multiclass Problems , 2002, Machine Learning.

[2]  Shin Ishii,et al.  A Unified Framework of Binary Classifiers Ensemble for Multi-class Classification , 2012, ICONIP.

[3]  Thomas Phan,et al.  Improving activity recognition via automatic decision tree pruning , 2014, UbiComp Adjunct.

[4]  Gamze Uslu,et al.  RAM: Real Time Activity Monitoring with feature extractive training , 2015, Expert Syst. Appl..

[5]  B. Fei,et al.  Binary tree of SVM: a new fast multiclass training and classification algorithm , 2006, IEEE Transactions on Neural Networks.

[6]  Tae-Seong Kim,et al.  A Triaxial Accelerometer-Based Physical-Activity Recognition via Augmented-Signal Features and a Hierarchical Recognizer , 2010, IEEE Transactions on Information Technology in Biomedicine.

[7]  B. Zadrozny.  Reducing multiclass to binary by coupling probability estimates , 2001, NIPS.

[8]  Yongqiang Zhang,et al.  Regression Cloud Models and Their Applications in Energy Consumption of Data Center , 2015, J. Electr. Comput. Eng..

[9]  Dong Wang,et al.  Learning machines: Rationale and application in ground-level ozone prediction , 2014, Appl. Soft Comput..

[10]  Binh Q. Tran,et al.  Optimization of an Accelerometer and Gyroscope-Based Fall Detection Algorithm , 2015, J. Sensors.

[11]  Daniel Gillblad,et al.  Learning Machines , 2020, AAAI Spring Symposia.

[12]  Shin Ishii,et al.  Ternary Bradley-Terry model-based decoding for multi-class classification and its extensions , 2011, 2008 IEEE Workshop on Machine Learning for Signal Processing.

[13]  Paul J. M. Havinga,et al.  Fusion of Smartphone Motion Sensors for Physical Activity Recognition , 2014, Sensors.

[14]  Ana M. Bernardos,et al.  Activity logging using lightweight classification techniques in mobile devices , 2012, Personal and Ubiquitous Computing.

[15]  Sergio Escalera,et al.  Boosted Landmarks of Contextual Descriptors and Forest-ECOC: A novel framework to detect and classify objects in cluttered scenes , 2007, Pattern Recognit. Lett..

[16]  Kenji Mase,et al.  Activity and Location Recognition Using Wearable Sensors , 2002, IEEE Pervasive Comput..

[17]  David M. Magerman.  Statistical Decision-Tree Models for Parsing , 1995, ACL.

[18]  Thomas G. Dietterich,et al.  Solving Multiclass Learning Problems via Error-Correcting Output Codes , 1994, J. Artif. Intell. Res..

[19]  Paul J. M. Havinga,et al.  Towards Physical Activity Recognition Using Smartphone Sensors , 2013, 2013 IEEE 10th International Conference on Ubiquitous Intelligence and Computing and 2013 IEEE 10th International Conference on Autonomic and Trusted Computing.

[20]  Sung-Bae Cho,et al.  Human activity recognition with smartphone sensors using deep learning neural networks , 2016, Expert Syst. Appl..

[21]  Jordi Vitrià,et al.  Discriminant ECOC: a heuristic method for application dependent design of error correcting output codes , 2006, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[22]  Elif Derya íbeyli.  Wavelet/mixture of experts network structure for EEG signals classification , 2008 .

[23]  Héctor Pomares,et al.  Human activity recognition based on a sensor weighting hierarchical classifier , 2013, Soft Comput..

[24]  Y. Schutz,et al.  A new accelerometric method to assess the daily walking practice , 2002, International Journal of Obesity.

[25]  Faicel Chamroukhi,et al.  An Unsupervised Approach for Automatic Activity Recognition Based on Hidden Markov Model Regression , 2013, IEEE Transactions on Automation Science and Engineering.

[26]  Daphne Koller,et al.  Discriminative learning of relaxed hierarchy for large-scale visual recognition , 2011, 2011 International Conference on Computer Vision.

[27]  Olivier Chételat,et al.  Very Low Complexity Algorithm for Ambulatory Activity Classification , 2005 .

[28]  Hui Xue,et al.  Can under-exploited structure of original-classes help ECOC-based multi-class classification? , 2012, Neurocomputing.

[29]  Michael L. Littman,et al.  Activity Recognition from Accelerometer Data , 2005, AAAI.

[30]  Juan José Pantrigo,et al.  Human activity recognition based on kinematic features , 2014, Expert Syst. J. Knowl. Eng..

[31]  Anderson Rocha,et al.  Multiclass From Binary: Expanding One-Versus-All, One-Versus-One and ECOC-Based Approaches , 2014, IEEE Transactions on Neural Networks and Learning Systems.

[32]  Maja Pantic,et al.  A Dynamic Texture-Based Approach to Recognition of Facial Actions and Their Temporal Models , 2010, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[33]  Sergio Escalera,et al.  On the Decoding Process in Ternary Error-Correcting Output Codes , 2010, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[34]  Jake K. Aggarwal,et al.  Human activity recognition from 3D data: A review , 2014, Pattern Recognit. Lett..

[35]  Sergio Escalera,et al.  A genetic-based subspace analysis method for improving Error-Correcting Output Coding , 2013, Pattern Recognit..

[36]  Daniel P. Siewiorek,et al.  Activity recognition and monitoring using multiple sensors on different body positions , 2006, International Workshop on Wearable and Implantable Body Sensor Networks (BSN'06).

[37]  Héctor Pomares,et al.  Daily living activity recognition based on statistical feature quality group selection , 2012, Expert Syst. Appl..

[38]  Robert B. McGhee,et al.  Estimation of Human Foot Motion During Normal Walking Using Inertial and Magnetic Sensor Measurements , 2012, IEEE Transactions on Instrumentation and Measurement.

[39]  Marco Altini,et al.  Transfer Learning in Body Sensor Networks Using Ensembles of Randomized Trees , 2015, IEEE Internet of Things Journal.

[40]  Albrecht Schmidt,et al.  Multi-sensor Activity Context Detection for Wearable Computing , 2003, EUSAI.

[41]  Sergio Escalera,et al.  On the design of an ECOC-Compliant Genetic Algorithm , 2014, Pattern Recognit..

[42]  Koby Crammer,et al.  Multiclass classification with bandit feedback using adaptive regularization , 2012, Machine Learning.

[43]  Özlem Durmaz Incel,et al.  User, device and orientation independent human activity recognition on mobile phones: challenges and a proposal , 2013, UbiComp.

[44]  Chris D. Nugent,et al.  A Lightweight Hierarchical Activity Recognition Framework Using Smartphone Sensors , 2014, Sensors.

[45]  Chi-Woong Mun,et al.  Comparison of k-nearest neighbor, quadratic discriminant and linear discriminant analysis in classification of electromyogram signals based on the wrist-motion directions , 2011 .

[46]  Cordelia Schmid,et al.  Constructing Category Hierarchies for Visual Recognition , 2008, ECCV.

[47]  Héctor Pomares,et al.  mHealthDroid: A Novel Framework for Agile Development of Mobile Health Applications , 2014, IWAAL.

[48]  Sergio Escalera,et al.  Subclass Problem-Dependent Design for Error-Correcting Output Codes , 2008, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[49]  Lakshmish Ramaswamy,et al.  A Hierarchical Meta-Classifier for Human Activity Recognition , 2016, 2016 15th IEEE International Conference on Machine Learning and Applications (ICMLA).

[50]  Kanak Saxena,et al.  Performance Evaluation of SVM and K-Nearest Neighbor Algorithm over Medical Data set , 2012 .

[51]  Ning Jia,et al.  Decoding design based on posterior probabilities in Ternary Error-Correcting Output Codes , 2012, Pattern Recognit..

[52]  Tae-Seong Kim,et al.  A single tri-axial accelerometer-based real-time personal life log system capable of human activity recognition and exercise information generation , 2011, Personal and Ubiquitous Computing.

[53]  Kwang Ryel Ryu,et al.  Activity Recognition by Smartphone Accelerometer Data Using Ensemble Learning Methods , 2015 .

[54]  Yoram Singer,et al.  Reducing Multiclass to Binary: A Unifying Approach for Margin Classifiers , 2000, J. Mach. Learn. Res..

[55]  Richard Frisby,et al.  Comparison of Feature Classification Algorithm for Activity Recognition Based on Accelerometer and Heart Rate Data , 2009 .

[56]  Meng Li,et al.  A Model with Hierarchical Classifiers for Activity Recognition on Mobile Devices , 2016, 2016 IEEE Trustcom/BigDataSE/ISPA.

[57]  Ling Bao,et al.  Activity Recognition from User-Annotated Acceleration Data , 2004, Pervasive.

[58]  W. W. Peterson,et al.  Error-Correcting Codes. , 1962 .

[59]  Chih-Jen Lin,et al.  Probability Estimates for Multi-class Classification by Pairwise Coupling , 2003, J. Mach. Learn. Res..

[60]  Sergio Escalera,et al.  An incremental node embedding technique for error correcting output codes , 2008, Pattern Recognit..

[61]  Yuhuang Zheng,et al.  Human Activity Recognition Based on the Hierarchical Feature Selection and Classification Framework , 2015, J. Electr. Comput. Eng..

[62]  Huiru Zheng,et al.  Activity Monitoring Using a Smart Phone's Accelerometer with Hierarchical Classification , 2010, 2010 Sixth International Conference on Intelligent Environments.

[63]  Hakob Sarukhanyan,et al.  ACTIVITY RECOGNITION USING K-NEAREST NEIGHBOR ALGORITHM ON SMARTPHONE WITH TRI-AXIAL ACCELEROMETER , 2012 .

[64]  Davide Anguita,et al.  Human Activity Recognition on Smartphones Using a Multiclass Hardware-Friendly Support Vector Machine , 2012, IWAAL.

[65]  Tadahiro Kuroda,et al.  Wearable sensor-based human activity recognition from environmental background sounds , 2012, Journal of Ambient Intelligence and Humanized Computing.

[66]  J Lubitz,et al.  The effect of longevity on spending for acute and long-term care. , 2000, The New England journal of medicine.

[67]  Elif Derya Übeyli.  Wavelet/mixture of experts network structure for EEG signals classification , 2008, Expert Syst. Appl..

[68]  Mita Nasipuri,et al.  Performance Comparison of SVM and ANN for Handwritten Devnagari Character Recognition , 2010, ArXiv.

[69]  Bernardo Nugroho Yahya,et al.  AN EFFECTIVE THRESHOLD BASED MEASUREMENT TECHNIQUE FOR FALL DETECTION USING SMART DEVICES , 2017 .

[70]  Nima Hatami,et al.  Thinned-ECOC ensemble based on sequential code shrinking , 2012, Expert Syst. Appl..

[71]  Robert Tibshirani,et al.  Classification by Pairwise Coupling , 1997, NIPS.

[72]  Martin T. Hagan,et al.  Neural network design , 1995 .

[73]  Kaizhu Huang,et al.  Joint learning of error-correcting output codes and dichotomizers from data , 2011, Neural Computing and Applications.

[74]  Cagatay Catal,et al.  On the use of ensemble of classifiers for accelerometer-based activity recognition , 2015, Appl. Soft Comput..

[75]  Wei Hu,et al.  AdaBoost-Based Algorithm for Network Intrusion Detection , 2008, IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics).

[76]  Moustapha Cissé,et al.  Learning Compact Class Codes for Fast Inference in Large Multi Class Classification , 2012, ECML/PKDD.

[77]  Yuting Zhang,et al.  Continuous functional activity monitoring based on wearable tri-axial accelerometer and gyroscope , 2011, 2011 5th International Conference on Pervasive Computing Technologies for Healthcare (PervasiveHealth) and Workshops.

[78]  F. Lemmermeyer.  Error-correcting Codes , 2005 .

[79]  Hongnian Yu,et al.  Elderly activities recognition and classification for applications in assisted living , 2013, Expert Syst. Appl..