Dealing with Imbalanced Data Sets for Human Activity Recognition Using Mobile Phone Sensors

In the recent years, the wide spreading of smart-phones which are daily carried by humans and fit with tens of sensors triggered an intense research activity in human activity recognition (HAR). HAR in smartphones is seen as essential not only to better understand human behavior in daily life but also for context provision to other applications in the smartphone. Many statistical and logical based models for on-line or off-line HAR have been designed, however, the current trend is to use deep-learning with neural network. These models need a high amount of data and, as most discriminative models, they are very sensitive to the imbalanced class problem. In this paper, we study different ways to deal with imbalanced data sets to improve accuracy of HAR with neu-ral networks and introduce a new oversampling method, called Border Limited Link SMOTE (BLL SMOTE) which improves the classification accuracy of Multi-Layer Perceptron (MLP) performances.

[1]  Michel Vacher,et al.  On-line human activity recognition from audio and home automation sensors: Comparison of sequential and non-sequential models in realistic Smart Homes , 2016, J. Ambient Intell. Smart Environ..

[2]  Hui Han,et al.  Borderline-SMOTE: A New Over-Sampling Method in Imbalanced Data Sets Learning , 2005, ICIC.

[3]  Miguel A. Labrador,et al.  A Survey on Human Activity Recognition using Wearable Sensors , 2013, IEEE Communications Surveys & Tutorials.

[4]  Weihua Sheng,et al.  Multi-sensor fusion for human daily activity recognition in robot-assisted living , 2009, 2009 4th ACM/IEEE International Conference on Human-Robot Interaction (HRI).

[5]  Damla Arifoglu,et al.  Activity Recognition and Abnormal Behaviour Detection with Recurrent Neural Networks , 2017, FNC/MobiSPC.

[6]  Haibo He,et al.  Learning from Imbalanced Data , 2009, IEEE Transactions on Knowledge and Data Engineering.

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

[8]  Gert R. G. Lanckriet,et al.  Recognizing Detailed Human Context in the Wild from Smartphones and Smartwatches , 2016, IEEE Pervasive Computing.

[9]  C. Lee Giles,et al.  Learning on the border: active learning in imbalanced data classification , 2007, CIKM '07.

[10]  Seyda Ertekin,et al.  Adaptive Oversampling for Imbalanced Data Classification , 2013, ISCIS.

[11]  Sihem Amer-Yahia,et al.  Scalable Active Temporal Constrained Clustering , 2018, EDBT.

[12]  Gary M. Weiss,et al.  Activity recognition using cell phone accelerometers , 2011, SKDD.

[13]  Duc A. Tran,et al.  The 11th International Conference on Mobile Systems and Pervasive Computing (MobiSPC-2014) A Study on Human Activity Recognition Using Accelerometer Data from Smartphones , 2014 .

[14]  Burr Settles,et al.  Active Learning Literature Survey , 2009 .

[15]  David W. Jacobs,et al.  Active image clustering: Seeking constraints from humans to complement algorithms , 2012, 2012 IEEE Conference on Computer Vision and Pattern Recognition.

[16]  Nitesh V. Chawla,et al.  SMOTE: Synthetic Minority Over-sampling Technique , 2002, J. Artif. Intell. Res..

[17]  Ira Assent,et al.  AnyDBC: An Efficient Anytime Density-based Clustering Algorithm for Very Large Complex Datasets , 2016, KDD.

[18]  Doruk Coskun,et al.  On-line Context Aware Physical Activity Recognition from the Accelerometer and Audio Sensors of Smartphones , 2014, AmI.

[19]  Miguel A. Labrador,et al.  A mobile platform for real-time human activity recognition , 2012, 2012 IEEE Consumer Communications and Networking Conference (CCNC).

[20]  Belkacem Fergani,et al.  A New Multi-Class WSVM Classification to Imbalanced Human Activity Dataset , 2014, J. Comput..

[21]  Sang Joon Kim,et al.  A Mathematical Theory of Communication , 2006 .