Evaluation of Clustering Algorithms on GPU-Based Edge Computing Platforms

. 2020 Nov 6;20(21):6335. doi: 10.3390/s20216335

Algorithm 1 FCM algorithm in GPU.

1:
$i n i t_u < < > > (s t a t e s, f u z z y_u, c l u s t e r s, r o w s);$
2:
for $s t e p = 1; i < M A X_S T E P S; i = i + 1$ do
3:
$c u b l a s D g e m m_c o n f i g u r a t i o n_{u} (f u z z y_u, m f, v e c t o r_u_t r a n s f o r m, c l u s t e r s, r o w s, c o l u m n s);$
4:
$n u m e r a t o r_C e n t r o i d_{i} < < > > (n u m e r a t o r, m f, c o l u m n s, r o w s, c l u s t e r s)$
5:
$d e t e r m i n a t o r_C e n t r o i d_{i} < < > > (d e n o m i n a t o r, v e c t o r_u_t r a n s f o r m, c o l u m n s, r o w s, c l u s t e r s)$
6:
$o b t a i n_c e n t e r s < < > > (c e n t e r, n u m e r a t o r, d e n o m i n a t o r, m f, c l u s t e r s, r o w s, c o l u m n s)$
7:
$d i s t a n c e_m a t r i x < < > > (d i s t a n c e, c e n t e r, d a t a_{m} a t r i x, r o w s, c o l u m n s)$
8:
$E r r o r_s t e p_{i} = t h r u s t : : r e d u c e (m f);$
9:
if $(E r r o r_s t e p_{i} - E r r o r_s t e p_(i - 1)) < e r r o r$ then
10:
$b r e a k;$
11:
end if
12:
$n e w_u_f o r_n e x t_i t e r a t i o n < < > > (f u z z y_u, d i s t a n c e, r o w s, c o l u m n s, c l u s t e r s);$
13:
end for
14:
$c u d a M e m c p y (u_h o s t, f u z z y_u, c l u s t e r * r o w s * s i z e o f (F L O A T), c u d a M e m c p y D e v i c e T o H o s t);$