…so the PinSAGE style neighbourhood sampling method dgl.sampling.PinSAGESampler! This is basically how we could solve the training on large graphs through DGL new version. ▼【Hands-on guidance to DGL library】Series buildersbox.corp-sansan.com

DSOC研究員の吉村です. 最近はコンビニスイーツにハマってます. この時期 (執筆時点で10月ごろ) になると, 芋, 南瓜, 栗系の物を探しては色々試しています.さて, 前回からしばらく空いてしまいましたが今回は「グラフ」についてです. 近年では GCN や, GraphSAGE などのグラフ学習手法が有名ですが, 今回はもう少し別の観点から「グラフ」と機械学習について見ていきたいと思います.

Hi, I am XING LI, a researcher from Sansan DSOC. We have discussed some of the most important elements in DGL, such as Message Passing and NodeFlow. And we also explored a small but vital topic in most deep neural networks: (Negative) Sampl…

…topic of "Hands-on guidance to DGL library", it is more like my personal summary on graph edge negative sampling problem. No need to talk about the universality and importance of graph edge negative sampling, you could find some useful insi…

Hi, I am XING LI, a researcher from Sansan DSOC. Last time, we briefly discussed the DGL(Deep Graph Library) and one of its fundamental mechanisms: Message Passing. Today, we are going to explore another significant data structure: NodeFlow…

… a series hands-on guidance to DGL library together with graph learning topics as examples. This is the first blog of this series, starting with a brief introduction and one core feature of DGL. Introduction In the last few years, many chal…