Nicht nur am letzten item Loss berechnen, sondern an jedem Item Loss berechnen in RNN

metzzo / deep_recsys

Achieves 0.55 MRR (and 0.64 MRR on the specific subset of data) at the RecSys2019 challenge using a two stage deep neural network architecture. More information can be gathered from the Documentation.pdf in particular at the "Deep Learning Model" section..

http://programming-with-design.at

2 stars 0 forks source link

Nicht nur am letzten item Loss berechnen, sondern an jedem Item Loss berechnen in RNN #9

Closed metzzo closed 5 years ago

metzzo commented 5 years ago

keine verbesserung