yuflo / g-SNE-model

Graph Stochastic Neighbor Embedding

g-SNE-model

Graph Stochastic Neighbor Embedding

Introduction

This algorithm is partly inspired by LargVis and can be used to visualize large-scale homogeneous/heterogeneous information network.

Run

For homogeneous information network:

./gsne_hom -train data_directory -output emb.txt -binary 1 -size 2 -negative 5 -samples 10 -rho 0.025 -threads 10

For heterogeneous information network:

./gsne_hin -train data_directory -output emb.txt -binary 1 -size 2 -negative 5 -samples 10 -rho 0.025 -threads 10

• -train, the input file of network data;
• -output, the output file of the embedding;
• -binary, save the learnt embeddings in binary moded; default is 0 (off);
• -size, set dimension of vertex embeddings, default is 18;
• -negative, the number of negative samples used in negative sampling; the deault is 5;
• -samples, set the number of training samples as Million; default is 1;
• -threads, the total number of threads used; the default is 1.
• -rho, the starting value of the learning rate; the default is 0.025;
• -gamma, set the gamma, default is 7

Contact:

If you have any questions about the codes and data, please feel free to contact us.

Chen Li, [e-mail]