FM(factorization Machines)

《Factorization Machines》paper的閱讀筆記,僅爲了整理個人思路。

個人覺得FM的本質就是預測值=偏置+權重1單變量+權重2變量之間的相互作用。
偏置和權重都可以是標量,也可以是向量

下面是本人認爲重要的文章內容摘抄與翻譯,能力有限,水平不足,不喜請繞道。

一 FM的優點

- 能夠估計SVM所不能的稀疏矩陣的參數
(FMs allow parameter estimation under very sparse data where SVMs fail)

- FM 具有線性複雜性(相當於SVM中的多項核),能夠在原始數據中進行優化,無需像SVM一樣依賴支持向量。
(FMs have linear complexity,can be optimized in the primal and do not rely on support vectors like SVMs)

- FM 具有一般性,能夠適用於任何真實值的特徵向量,能夠模擬偏置MF,SVD++,PITF,FPMC等最先進的模型。
(FMs are a general predictor that can work with any real valued feature vector.In contrast to this ,other state-of-the-art factorization models work only on very restricted input data.We will show that just by defining the feature vectors of the input data,FMs can mimic state-of-the- art models like biased MF ,SVD++,PITF,or FPMC.)

二 FM模型的公式

y^(x)=w0+i=1nwixi+i1nj=i+1n<vi,vj>xixj\hat{y}(x) = w_0 +\sum_{i=1}^{n}w_ix_i + \sum_{i-1}^{n}\sum_{j=i+1}{n}<v_i,v_j>x_ix_j
w0Rw_0 \in R,wRnw\in R^n,VRnkV\in R^{n*k},
<.,.>是大小爲K的兩個向量的點積,&lt;vi,vj&gt;=f=1kvi,f.vj,f&lt;v_i,v_j&gt; = \sum_{f=1}^{k}v_{i,f}.v_{j,f}

V中的行向量viv_i代表的是有K個因子的第i個變量。
kN0+k \in N_{0}^{+} 是定義因子的超參。

(A row wiw_i within V describes the ii-th variable with k factors.kN0+k \in N_{0}^{+} is a hyperparameter that defines the dimensionality of the factorization)

自由度爲2的FM能夠捕捉單變量和變量之間相互作用。
(A 2-way FM(degree d = 2) captures all single and pairwise interactions between variables)

  • w0w_0 是全局變量
  • wiw_i模擬第i個變量的strength(個人覺得其實就是權重,models the strength of the i-th variable)
  • w^i,j=&lt;vi,vj&gt;\hat w_{i,j} = &lt;v_i,v_j&gt;模擬第i和第j個變量之間的相互作用。(個人覺得其實就是權重,models the interaction between the i-th and j-th variable)

三 FM模型的表達能力

假設K足夠大,對於任何正定矩陣W,存在一個矩陣V滿足V.VtV.V_t。也就是說,如果K的選擇足夠大,FM便能夠表達任意的相互作用向量W。爲了使模型具有更好的泛化能力,在稀疏數據集中,通常選在比較小的K,。

(It is well known that for any positive definite matrix W, there exists a matrix V such thta W=V.VtW=V.V^t provided that kk is large enough. Nevertheless, in sparse settings,typically a small kk shold be chosen because there is not engough data to estimate complex interactions W.Restricting K - and thus the expressiveness of the FM -leads to better generalization and thus improved interaction matrics under sparsity)

發表評論
所有評論
還沒有人評論,想成為第一個評論的人麼? 請在上方評論欄輸入並且點擊發布.
相關文章