Master-thesis
_Isomorphism
___Trees
_____Rootedtree.py
_Documents
___graphKernel
_____AaltoLecture-LearningOnGraphs
_____A short tour of Kernel method for Graphs (Thomas Garther (Germany) - Quoc V.Le, Alex J Smola (Austria))
_____Graph Kernel 2014 (Vishwanathan (Pordue) - Schraudolph (Australia) - Kondor (Pasadena) - Borgwardt (Germany))
_____NSPDK (Fabrizio Costa)
___garphNeuralNetwork
_____How powerful are GNN - (Jure Leskovec)
_____Graph Neural Networks with Convolutional ARMA Filters
___Autoencoder and network(Aalto)
___Statistical Comparisons of Classifiers(Janez Demasar)
__TestDataset
___AIDS (dataset folder)
___DD (dataset folder)
___KKI (dataset folder)
___MIO (dataset folder)
___MUTAG (dataset folder)
___PROTEINS (dataset folder)
___Import_data.ipynb (Notebook)
___K-mean_Evaluation.ipynb (Notebook)
___KKN_Evaluation.ipynb (Notebook)
___KKN_UMAP_Evaluation.ipynb (Notebook)
Is a python method that convert a Tree in a binary string, the idea comes from the book "Application of Graph Theory" of Robin J.Wilson and Lowell W.Beineke.
The algorithm generate the code starting from the root of the tree, if the tree has no root, the center is used as root.
It contains a method to import data. The format data comes from Dourmund Universitat.
The method produce a Networkx OBJ
It import the data, vectorize it using NSPDK, successively it use a SVD dimensionality reduction and finally it uses K-mean as classifier
It import the data, vectorize it using NSPDK, successively it use a SVD dimensionality reduction and finally it uses K-nearest neighbor as classifier
It import the data, vectorize it using NSPDK, successively it use a UMAP dimensionality reduction and finally it uses K-nearest neighbor as classifier