Title:
Developing an LSTM Pipeline for Accelerometer Data

Description:
We explore the architecture space available for the construction of deep LSTM (d-LSTM) recurrent neural network (RNN) models, specifically related to predictive analytics in the domain of human activity recognition (HAR) from accelerometer time series data.



Presenter(s):
 Speaker: Christian McDaniel, University of Georgia
 Speaker: Shannon Quinn, University of Georgia