ncTL {FGSG} | R Documentation |
Given A = {a_1,…,a_n}, the response y, and a set of edges E, this function aims to solves
min 1/2||Ax-y||^2 + λ_1 ∑\limits_i(J_τ(|x_i|)) + λ_2 ∑_{(i,j)\in E} w_(i,j)|x_i-r_{(i,j)}x_j|
where J_τ(x)=min(x/τ,1) is a surrogate of the L0 norm, w_(i,j) is the weight of the edge (i,j), and r_(i,j) is the sign of the correlation between features a_i and a_j. The weight and sign can be specified in Rwt: w=|Rwt|, and r=sign(Rwt).
ncTL(A, y, tp, s1, s2, RmaxIter = 100, RaMaxIter = 1000, Rrho = 5, Rtau = 0.15, Rwt = rep(1, length(tp)), Rtol = 0.001, RaTol = 0.001, x0 = rep(0, ncol(A)))
A |
|
y |
|
tp |
|
s1 |
|
s2 |
|
RmaxIter |
|
RaMaxIter |
|
Rrho |
|
Rtau |
|
Rwt |
|
Rtol |
|
RaTol |
|
x0 |
|
Returned value x0 is the solution to the optimizaiton problem.
Yiwen Sun
S.Yang, L.Yuan, Y.Lai, X.Shen, P.Wonka, and J.Ye. Feature grouping and selection over an undirected graph. KDD, 2012.
A<-matrix(rnorm(25),5,5) y<-rnorm(5) tp<-c(1,2,2,3,3,4,4,5) ncTL(A,y,tp,0,0)