Graph Neural Networks (GNNs) and Visualizations
Tags:
Status:
Active, Closed for joining
Contributor/Mentors
Contributor: Pakhi Banchalia
Mentors: Bradly Alicea, Himanshu Chougule
About
DevoGraph is an open-source GNN framework designed specifically for analyzing embryogenetic data. It utilizes Graph Neural Networks (GNNs) as a powerful method for revealing connectivity patterns within complex geometric datasets. My contribution to this project involves (i) implementing the Mapper algorithm for topological data analysis, (ii) implementing an Evolutionary-based Neural Developmental Program to mirror the embryogenetic growth of C. elegans, and (iii) code integration in DevoLearn. Through this project, I aim to create informative and interactive visualizations for better interpretation and analysis of the data.
Deliverables
2024
- Create informative and interactive visualizations for better interpretation and analysis of the data
2024