alibaba / alibaba/euler

请问当前版本是否支持关于边的监督学习?

Open
#120 2 comments 0 reactions 0 assignees View on GitHub
Dominant language
C++
Stars
2.9k
Forks
553
PR merge metrics
No merged PRs in 30d

Description

阿里研发团队的大佬们辛苦了。

1)在user manual一节中--label_idx在ppi数据集中是在node的float feature中,请问边的feature和点的feature是否可以一并顺序编号?

2)如果1)可以的话,是否可以指定--label_idx为边上的float_feature,从而进行边的有监督学习?(例如类似link prediction的task)

3)如果2)可以的话,请问正负采样的具体方法和比例,以及最终边的embedding?

其实就是对于--feature_idx,--feature_dim,--label_idx,--label_dim有些模糊,对数据中点边feature的构成也有些不太清晰。(也有一个issue问了这四个参数的意义,已参考)

谢谢大佬们~
====================
补充:
4)看Preparing Data里的例子,似乎点,边的feature不能顺序编号?上述四个参数只能特指node中的float feature? 甚至在点或者边内部的三个int,float和binay也是分别从0编号?

Contributor guide

No contributing guide indexed for this repository

Research direction

Start with the user manual and the Preparing Data example mentioned in the issue, then trace the definitions of feature_idx, feature_dim, label_idx, and label_dim. Determine whether edge features and labels, supervised edge learning, negative sampling, and edge embeddings are supported. Done requires a clear documented answer or a scoped design for the missing capability.

Written by the indexing model from the issue text.

Assessment

Domain
machine-learning
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
20/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.