Start from scratch
f = x1 v x2.x3 (Assume x2x3 is multipplication and not logical "and"?
My vectors that is going tobe used below:
==========================================
Use vectors is in the question is the threshold value always 1?
Why they tell use here threshold = 0in the question. Should my last
vector value not be 1 but 0? Is mine below incorrect
Vectors:
========
vectorx1 = <0 1 0 1>
vectorx2 = <1 0 0 1>
vectorx3 = <0 0 1 1>
vectorx4 = <1 1 1 1>
Training:
=========
Step 1:
=======
W1 = (0 0 0 0)
positional then x1 x2 x3 x4 match below to value (positional)
Vectorx1 = (0 1 0 1 )
so when calc d = x1 v x2x3 values in order of the X vector are used
so d = 0 v 1.0 = 0 v 0 = 0
W1.X = (0.0 + 0.1 + 0.0 + 0.1) = 0
So f = 1
Step 2:
=======
W2 = W1 + c(d-f)X1 = (0 0 0 0) + 1(0-1)(0 1 0 1 )
= (0 -1 0 -1)
positional then x1 x2 x3 x4 match below to value
Vector X = (1 0 0 1)
so when calc d = x1 v x2x3 values in order of the X vector are used
so d = 1 v 0.0 = 1 v 0 = 1
W2.X = (0 -1 0 -1).(1 0 0 1) = (0.1 + -1.0 + 0.0 + -1.1) = -1
So f = 0
Step 3:
=======
W3 = W2 + c(d-f)X2 = (0 -1 0 -1) + 1(1-0)(1 0 0 1)
= (0 -1 0 -1) + (1 0 0 1)
= (1 -1 0 0)
positional then x1 x2 x3 x4 match below to value
Vector X = (0 0 1 1)
so when calc d = x1 v x2x3 values in order of the X vector are used
so d = 0 v 0.1 = 0 v 0 = 0
W3.X = (1 -1 0 0).(0 0 1 1)=(1.0 + -1.0 + 0.1 + 0.1) = 0
So f = 1
Step 4:
=======
W4 = W3 + c(d-f)X3 = (1 -1 0 0) + 1(0-1)(0 0 1 1)
= (1 -1 0 0)+(0 0 -1 -1)
= (1 -1 -1 -1)
positional then x1 x2 x3 x4 match below to value
Vector X = (1 1 1 1)
so when calc d = x1 v x2x3 values in order of the X vector are used
so d = 1 v 1.1 = 1 v 1 = 1
W4.X = (1 -1 -1 -1).(1 1 1 1)=(1.1 + -1.1 + -1.1 + -1.1) = -2
So f = 0
Step 5:
=======
W5 = W4 + c(d-f)X4 = (1 -1 -1 -1) + 1(1-0).(1 1 1 1)
= (1 -1 -1 -1)+(1 1 1 1)
= (2 0 0 0)
positional then x1 x2 x3 x4 match below to value
Vector X = (1 0 1 1)
so when calc d = x1 v x2x3 values in order of the X vector are used
so d = 1 v 0.1 = 1 v 0 = 1
W4.X = (2 0 0 0).(1 0 1 1)=(2.1 + 0.0 + 0.1 + 0.1) = 2
So f = 1
Know that both f and d = 1 is this then final vector???
weight vector = w5 = (2 0 0 0) and threshold = 0
How do one know when to stop training??
Step 6:
=======
W6 = W5 + c(d-f)X5 = (2 0 0 0) + 1(1-1).(0 1 1 0)
= (2 0 0 0)
If w6 = w5 then w5 is last weight vector???
Is my answer correct or must I use threshold = 0
in vectors as in q1.4 states threshold = 0
So all my input vectors must have 0 instead of
1 at end ??????
Vectors:
========
vectorx1 = <0 1 0 0> I used above vector <0 1 0 1>
vectorx2 = <1 0 0 0> I used above vector <1 0 0 1>
vectorx3 = <0 0 1 0> I used above vector <0 0 1 1>
vectorx4 = <1 1 1 0> I used above vector <1 1 1 1>
