Skip to content

shellrazer/CTR_tools

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 
 
 

Repository files navigation

CTR_tools

This repository records some useful tools in machine learning of recommendation.

  1. add bayesian smoothing class. I use moment method in sql calculation of conversion rate to solve 1) empty rate 2) ambiguous rate, eg: 1/1 has a higher rate than 9/10.

  2. add common attention layers. I use additive attention in deepfm to give model more flexibility on regions. regions are different cluster of people. yes we do not train seperated model on different cluster of people.

    Implementation list

    Name Alignment score function Citation
    Additive score(s_t, h_i) = v tanh(W[s_t;h_i]) Bahdanau 2015
    Dot-Product score(s_t, h_i) = s_t · h_i Luong 2015
    Scaled Dot-Product score(s_t, h_i) = s_t · h_i / d_k Vaswani 2017

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages