Experimental and Analysis of Ensemble Deep Learning for Fall Detection using a smartwatch

Abstract

In this talk, I will present an experimental study of Ensemble Deep Learning (DL) techniques for the analysis of time series data sensed by smart watches, DL has demonstrated superior performance as compared to traditional machine learning techniques on fall detection applications due to the fact that important features in time series data can be learned and need not be determined manually by the domain expert. However, DL networks generally require large datasets for training. In fall detection, there are no publicly available large annotated datasets that can be used for training, due to the nature of the problem (i.e. a fall is a rare event). Moreover, fall data is also inherently noisy since many motions generated by the wrist-worn smartwatch can be mistaken for a fall. We explore combing DL (Recurrent Neural Network) with ensemble techniques (Stacking and AdaBoosting.. We conducted a series of experiments using two different datasets of simulated falls for training various ensemble models. Our results show that an ensemble of deep learning models combined by the stacking ensemble technique, outperforms a single deep learning model trained on the same data samples, and thus, may be better suited for small-size datasets.

In this second part of the talk, I will briefly describe the unique nature of the Computer Science PhD program at Texas State University and the opportunities for funded PhD positions.

Bio

Anne H.H. Ngu is currently a Professor and the PhD Program Director with the Department of Computer Science at Texas State University. From 1992-2000, she worked as a Senior Lecturer in the School of Computer Science and Engineering, University of New South Wales, Australia.  She had held the research scientist/scholar position with Telcordia Technologies; Lawrence Livermore National Laboratory, Microelectonics and Computer Technology (MCC); University of California, Berkeley; Commonwealth Scientific and Industrial Research Organization (CSIRO), Australia and the Tilburg University, The Netherlands. Dr Ngu has published over 130 technical papers in journals and refereed conferences in computer science. Her main research interests are in large-scale service and information discovery and integration, Internet of Things software platform and applications, Scientific workflows, Databases and Software Engineering. Her professional service features key leadership roles in International conference on Data Engineering (ICDE)  and Web Information  Systems Engineering  conference (WISE).  She was a winner of the 2013 NCWIT Undergraduate Research Mentoring Award. 

Time and Location

February 22, 2022 @ 1:30PM, ZOOM https://sjsu.zoom.us/j/86925918978?pwd=U1V6UDNnNi9jaDFMU2dDZS92bTNpUT09