NA-MIC Project Weeks

Back to Projects List

Multivariate Functional Shape Data Analysis & Distributed Learning

Key Investigators

• Juan-Carlos Prieto (University of North Carolina)
• Loïc Michoud (University of Michigan)

Project Description

Objective

1. Develop a system to allow users of the website to use a statistical model (MFSDA)
2. Develop systems to allow users to use deep learning models about bony changes and disease changing in TMJ OA

Approach and Plan

1. Use of clusterpost that handles the data transfer, the execution and the monitoring of the job
2. We will add the deep learning models in the web using clusterpost

Progress and Next Steps

• Developing the preprocessing for MFSDA
• Adding the preprocessing as a plugin on the web site
• Adding the new models of deep learning in the web system
• Developing distributed learning to increase the accuracy of the models

Background and References

• Source Code
• Web Link
• Documentation