PageRankNibble_reimplementation

This is the student project in Rome-Moscow school 2020.

Requirements

Following packages can be installed using pip or conda.

networkx 2.5
numpy 1.19.2
pydot 1.3.0
pygraphviz 1.3
matplotlib 3.3.2
scipy 3.3.2
scikit-learn 0.23.2

Usage

python main_case1.py

Note that

• PageRankNibble algorithem is reimplemented in the function PageRankNibble_undirected.py.

• main_case1.py and main_case2.py use synthetic data.

• main_case3.py use UCI ML Wine recognition datasets. https://archive.ics.uci.edu/ml/machine-learning-databases/wine/wine.data

• To apply different k in k-NN algorithem, please modify the variable n_neighbors, and get accuracy under different k.

• Also you can choose a start point by modify the variable Seed.

Results

Experiment result on synthetic data (1/2)

Experiment result on synthetic data (2/2)

Experiment result on UCI ML Wine recognition datasets