search

hexiangnan / attentional_factorization_machine

TensorFlow Implementation of Attentional Factorization Machine
407 stars 156 forks source link

实数特征支持问题 #24

Open starspringcloud opened 3 years ago

starspringcloud commented 3 years ago

hi 拜读了你的代码，似乎不支持实值特征，要求特征值 为0/1， 但fm 理论上是支持实值特征的， 请问这么设计的原因是什么？

hexiangnan commented 3 years ago

想用fm支持实数值特征的话稍微改一下embedding层就行了。但这种方式处理实数值效果不好。建议先把实数值离散化

On Tue, Sep 22, 2020 at 9:40 PM starspringcloud notifications@github.com wrote:

hi 拜读了你的代码，似乎不支持实值特征，要求特征值 为0/1， 但fm 理论上是支持实值特征的， 请问这么设计的原因是什么？

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/hexiangnan/attentional_factorization_machine/issues/24, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAI3DDUI5CQR33DPFCOE3Z3SHCSL5ANCNFSM4RVZKGZQ .

-- Xiangnan HE (Dr) :: Professor :: Vice Dean, School of Data Science :: University of Science and Technology of China (USTC) :: 443 Huangshan Road, Hefei, China 230027 :: (0551) 63607236 (DID) :: hexn@ustc.edu.cn (E) :: http://staff.ustc.edu.cn/~hexn/ (W)