Fast Human Activity Recognition Based on a Massively Parallel Implementation of Random Forest

This article elaborates on the task of Human Activity Recognition being solved with the Random Forest algorithm. A performance measure is provided in terms of both recognition accuracy and computation speed. In addition, the Random Forest algorithm was implemented using CUDA, a technology providing options for massively parallel computations on low-cost hardware. The results suggest that Random Forest is a suitable and highly reliable technique for recognising human activities and that Graphics Processing Units can significantly improve the computation times of this otherwise rather time-consuming algorithm.

[1]  Alan R. Dabney.  BIOINFORMATICS Classification of Microarrays to Nearest Centroids , 2022 .

[2]  Mohamed Medhat Gaber,et al.  An Information-Theoretic Approach for Setting the Optimal Number of Decision Trees in Random Forests , 2013, 2013 IEEE International Conference on Systems, Man, and Cybernetics.

[3]  Paul Lukowicz,et al.  Collecting complex activity datasets in highly rich networked sensor environments , 2010, 2010 Seventh International Conference on Networked Sensing Systems (INSS).

[4]  Robert Tibshirani,et al.  Discriminant Adaptive Nearest Neighbor Classification , 1995, IEEE Trans. Pattern Anal. Mach. Intell..

[5]  Shyamsundar Rajaram,et al.  Human Activity Recognition Using Multidimensional Indexing , 2002, IEEE Trans. Pattern Anal. Mach. Intell..

[6]  Ricardo Chavarriaga,et al.  Benchmarking classification techniques using the Opportunity human activity dataset , 2011, 2011 IEEE International Conference on Systems, Man, and Cybernetics.

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

[8]  Thomas G. Dietterich.  Multiple Classifier Systems , 2000, Lecture Notes in Computer Science.

[9]  Leo Breiman,et al.  Random Forests , 2001, Machine Learning.

[10]  Misha Denil,et al.  Narrowing the Gap: Random Forests In Theory and In Practice , 2013, ICML.

[11]  Sotiris B. Kotsiantis,et al.  Bagging and boosting variants for handling classifications problems: a survey , 2013, The Knowledge Engineering Review.

[12]  Yoav Freund,et al.  An Adaptive Version of the Boost by Majority Algorithm , 1999, COLT '99.

[13]  Leo Breiman,et al.  Classification and Regression Trees , 1984 .

[14]  Hugo Fuks,et al.  Wearable Computing: Accelerometers' Data Classification of Body Postures and Movements , 2012, SBIA.

[15]  Wei-Yin Loh,et al.  Classification and regression trees , 2011, WIREs Data Mining Knowl. Discov..

[16]  James W. Davis,et al.  The Recognition of Human Movement Using Temporal Templates , 2001, IEEE Trans. Pattern Anal. Mach. Intell..

[17]  Didier Stricker,et al.  Creating and benchmarking a new dataset for physical activity monitoring , 2012, PETRA '12.

[18]  Bogdan Trawinski,et al.  Empirical Comparison of Bagging Ensembles Created Using Weak Learners for a Regression Problem , 2011, ACIIDS.

[19]  J. Ross Quinlan,et al.  C4.5: Programs for Machine Learning , 1992 .

[20]  Ying Wang,et al.  Human Activity Recognition Based on R Transform , 2007, 2007 IEEE Conference on Computer Vision and Pattern Recognition.