Cloud Based Infrastructure for Educational Deep Learning
Cameron T. Jones, Nicholas Anthony Diaz · Digital WPI · 2017
The aim of this project was to introduce deep learning functionality to the ASSISTments testbed. State of the art deep learning techniques require more advanced infrastructure, and ASSISTments has realized the need for such infrastructure to support these new methods of analysis for researchers. Previously researchers were able to get statistical analysis of their educational experiments through The Assessment for Learning Infrastructure (ALI), but were unable to take advantage of more advanced techniques. To correct this, our project implements a new workflow to streamline and strengthen the report generation process. We use a Python API hosted on Amazon cloud machines to achieve this goal, creating a powerful, yet simple and scalable infrastructure for future deep learning tasks.