Part of the NUS DBsystem Structured Data Project
ARMOR is a lightweight framework tailored for structured data (or relational data, tabular data) analytics.
Most businesses to date rely on structured data for their data storage and predictive analytics. Extracting insights from structured data requires advanced analytics.
ARMOR is mainly based on more effective and interpretable ARM-Net, an adaptive relation modeling network tailored for structured data.
View our ARM-Net publication in the ACM Special Interest Group on Management of Data (SIGMOD) 2021:
ARM-Net: Adaptive Relation Modeling Network for Structured Data
ARMOR source code is publicly available on GitHub.
Check out our ARMOR Project on Github
Explore our other repositories on Github
Extracting insights from structured data requires advanced analytics. The major challenge for the predictive modeling of structured data is how to model dependencies and correlations among features.
ARMOR focuses on relation modeling for structured data with state-of-the-art techniques. This framework is mainly based on more effective and interpretable ARM-Net.