graph-feature-extractor

This repository is a course project on Machine Learning course at Skoltech.

The project is a replication of work

Short description

Graph can be seen as a subset of the more general object named simplicial complex, which consists of elements of higher-orders (i.e. triangles, tetrahedras, etc.), in addition to nodes and edges. Practically, this allows to encode higher-order interactions, while graphs only encode pairwise ones. Concept of node centrality of graphs can be generalized to higher-order elements of simplical complexes. As a result, one could extract additional higher-order local information from a simplicial complex model compared to graph one. Project requires basic knowledge of graph theory and algorithms on graphs.

Files logistics

1. predicting.ipynb - file containing code for all experiments
2. graph_utils.py - file containing wrappers for feature extraction methods
3. simplcompl.py - file containing high level feature extraction described in the paper
4. project_presentation.pdf - pdf with presentation of the project
5. project_report.pdf - pdf with report of the project