Ve ctoY
Sußpovt achine
Euation to calulate weipht vectoys
vector machine .
Supsort
hyber ptone eqaion
Supþor t Ve ctoY rTachi nes
Numeri cal eLamþle
thee b eloug
the
hyper plan for
data po ints
th eue ae
eight data boints
To tal
Pogitively labeled data boints
4):(). (1).()}
Neg atively labeleol
labele data þointy
i(6).(). C).())
3+
o()
(4)
^/!)
-2 Thyee sUpbort vectoys =
Ss (?)
-3 G)
three dla ta þoint be ausmenteol
the
bias inhut .
So, S =
(0) ()
)
S
Three
(E)
variables
3-()
) 13
Und to
calelate
weight vehy
NO w
(S
+I(syas 4f positively
labellecL data set
-|
()4) +
()0): (OC)EX0)
9))-6X0)4)
byoduct +d vectoYs
dot poduct
fnd out
<(1+0+)+ (3t0)+ 3(3to+) =l
2
Arith meatic prreuion
=
By -35
factor w
Nouw need to (aleula e weith
) ) )
3 0-75
3
+ 07s
= -35
Hyher blane egvatiom
hyjerplone
b= -2 3+
( S
lne i parallel
-2 -!
-2
-3+