Table - PMC

Skip to main content

View full-text article in PMC

. 2017 Aug 8;17(8):1824. doi: 10.3390/s17081824

Algorithm 1. Smooth

l_{0}

-norm Algorithm

INPUT Data: W = SL0(A,x,

Σ_{m i n}

,

A_{p i n v}

);

OUTPUT Result: return weight vector W;

1:
$W = A_{p i n v} * x$ ;
2:
$σ = 2 * m a x (a b s (W))$ ;
3:
$μ_{0} = 2$ ;
4:
$L = 3$ ;
5:
$σ_{d e c r e a s e} = 0.5$ ;
6:
$Σ_{m i n} = 0.01$
7:
While ( $σ < Σ_{m i n}$ );
8:
For ( $i = 1 : L$ );
9:
$δ = W . * e x p (- a b s (W) .^{2} / σ^{2})$ ;
10:
$W = W - μ_{0} * δ$ ;
11:
$W = W - A_{p i n v} * (A * W - x)$ ;
12:
end for;
13:
$σ = σ * σ_{d e c r e a s e}$ ;
14:
end while;