Zhou Xin

ml-1m数据集结果不一致

15

Hi，thanks for sharing the code. With your source code unmodified (dropout: 0.5, neg-weight: 0.5), I have tried on ml-1m and get the following results: First col: Recall Second col: NDCG...

EHCF_Sin 在ML1M上的结果

7

Hi, 作者，您好。 感谢您分享代码，我用您的原始代码跑了ML1M的数据集（无改动），Test的结果是这样的： > nan nan > 0.0 0.0 > 0.0006718172657037286 0.00012831548886966677 不知道 您那边自己跑过没？看原代码核心就是对 Positive rating的User-Item Matrix进行拟合，虽然不进行negative sampling，但是这样理论上效果不会好于MF之类的。 算法的核心在公式（8）： > self.loss1 = self.weight1 * tf.reduce_sum( > tf.reduce_sum(tf.reduce_sum(tf.einsum('ab,ac->abc', self.iidW, self.iidW), 0)...

### Reference Issues _No response_ ### Summary Visualization~ ### Basic Example Such as the plot in "information panel"? [visualization](https://raw.githubusercontent.com/Cinnamon/kotaemon/main/docs/images/preview-graph.png) ### Drawbacks NA ### Additional information _No response_