Summary of Graph As a Feature: Improving Node Classification with Non-neural Graph-aware Logistic Regression, by Simon Delarue et al.
Graph as a feature: improving node classification with non-neural graph-aware logistic regression
by Simon Delarue, Thomas Bonald, Tiphaine Viard
First submitted to arxiv on: 19 Nov 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper introduces Graph-aware Logistic Regression (GLR), a non-neural model for node classification tasks. Unlike traditional graph algorithms and Graph Neural Networks (GNNs), which rely on message passing, GLR encodes each node’s relationships as an additional feature vector, combining it with the node’s self attributes. The proposed approach outperforms both foundational and sophisticated state-of-the-art GNN models in node classification tasks, while achieving significant gains in computation time up to two orders of magnitude compared to its best neural competitor. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper helps us better understand how computers can be taught to classify things on complex networks like social media or traffic patterns. Right now, computers struggle when the same type of thing is connected to other similar things. To solve this problem, scientists developed a new way to do node classification called Graph-aware Logistic Regression (GLR). Unlike other methods that only look at part of the information available, GLR uses both what’s special about each “node” and how all the nodes are connected. Tests show that GLR works better than other methods in some cases. |
Keywords
* Artificial intelligence * Classification * Gnn * Logistic regression
