Combining Factorization Model and Additive
Forest for Recommendation
Presenter: Tianqi Chen
Team ACMClass@SJTU
August 11, 2012
Team ACMClass@SJTU
I
Original team name: undergrads
I
Members are students from ACMClass in SJTU
I
All members are undergraduates, except the presenter:)
1/14 ACMClass@SJTU Combining Factorization Model and Additive Forest for Recommendation
Overview of Our Solution
social network/acon
user age/gender
item taxonomy
mestamp …
Factorizaon Models
Addive Forest
Final Soluon
Rank
Opmizaon
Incorporated Informaon
Modeling Approach
Combinaon
Focus point of this presentaon
One Joint Model, No Ensemble
2/14 ACMClass@SJTU Combining Factorization Model and Additive Forest for Recommendation
Feature-based Matrix Factorization
ˆr
ui
=
X
c∈C (u)
α
(u)
c
p
c
T
X
c∈C (i)
β
(i)
c
q
c
+
X
c∈C (u,i)
γ
(u,i)
c
g
c
(1)
I
Θ = {p, q, g }, trained via stochastic gradient descent
I
α
(u)
c
: user feature of user u: user social network/action,
keyword/tag
I
β
(i)
c
: item feature weight of item(celeberity) i: item
taxonomy/network
I
γ
(u,i)
c
: global feature related to interaction between u and i:
user age/gender bias
3/14 ACMClass@SJTU Combining Factorization Model and Additive Forest for Recommendation
- 1
- 2
前往页