Markov Blanket =
Pliska)? =L – Mbl- a)
IO, I
P(B|A) )
0.0 plbla T 0.4
0.9
children + parents + chiding Panay
P (A)
0.4
P(D)
0.6
A
A
A
P(C|A)
T F
0.5 0.9
F
BD
TT TF FT FF
P(E|B,D)
0.4 0.5 0.2 0.1
What’s the Markov blanket of B? c -Trenn, King’s College London
2
B
E
C
D
Prior Sampling
PCa ie , b , – e r d)
A P(B|A) T 0.4
0.9
B D P(E|B,D)
TT 0.4 TF 0.5 FT 0.2 FF 0.1
A P(C|A) T 0.5
F
B F 0.9 E
P (A) 0.4
A
–
pool p
. =P(a)- a)-
tida) d)
Calculate P pe, aq
random numbers: 0.2, 0.3, 0.4, 0.1, 0.5, 0.2, 0.5, 0.1, 0.3, 0.4, 0.9, 0.9, 0 , 0.1, 0.2
c -Trenn, King’s College London 3
P(D) 0.6
C D
•
P
b i d)
In ¢H e 1
Melais,> cird)
Prior Sampling
Pl a e b)=
ie is
cd)t(a ier + ad)
Plaeis P d), , a
, , ,
PCa e,b,
P (A)
0.4
+Pla, e, bpc,
i
d)
A
P(D)
0.6
A
P(C|A)
T F
0.5 0.9
C
D
A
T F
BD
TT TF FT FF
P(B|A)
0.4 0.9
P(E|B,D)
0.4 0.5 0.2 0.1
B
E
Calculate P pe, aq
random numbers: 0.2, 0.3, 0.4, 0.1, 0.5, 0.2, 0.5, 0.1, 0.3, 0.4, 0.9, 0.9, 0 , 0.1, 0.2
c -Trenn, King’s College London 3
Prior Sampling
→ ☒ is – ee , – diet ✓I →,e)
A P(B|A) T 0.4
F 0.9
B D P(E|B,D)
P(D) 0.6
C D
True pc.dk/–‘-z
if Rand← Proba set true
TT 0.4 TF 0.5 FT 0.2 FF 0.1
E
c -Trenn, King’s College London
✗✗ ✗
3
P(A) 0.4
→ Ma, is, acid,-
A ha,>5,7C,>d,-e)✗
A
B F 0.9
P(C|A) T 0.5 ,
ik¥e
False
CalculatePBpeo,oaqA0 C D EA8C DEAOCDE random numbers: 0.2, 0.3, 0.4, 0.1, 0.5, 0.2, 0.5, 0.1, 0.3, 0.4, 0.9, 0.9, 0 , 0.1, 0.2
a bed- a ibcd.ae Tabo de
—
Prior Sampling
Pldle )
?
P (A)
0.4
A
P(D)
0.6
A
P(C|A)
T F
0.5 0.9
C
D
A
T F
BD
TT TF FT FF
P(B|A)
0.4 0.9
P(E|B,D)
0.4 0.5 0.2 0.1
B
E
Calculate P pe, aq
random numbers: 0.2, 0.3, 0.4, 0.1, 0.5, 0.2, 0.5, 0.1, 0.3, 0.4, 0.9, 0.9, 0 , 0.1, 0.2
c -Trenn, King’s College London 3
Rejection Sampling
P (A)
0.4
A
P(D)
0.6
A
P(C|A)
T F
0.5 0.9
C
A
T F
BD
TT TF FT FF
P(B|A)
0.4 0.9
P(E|B,D)
0.4 0.5 0.2 0.1
B
Calculate P pa|e, dq
random numbers: 0.2, 0.3, 0.4, 0.1, 0.5, 0.2, 0.5, 0.1, 0.3, 0.4, 0.9, 0.9, 0 , 0.1, 0.2
c -Trenn, King’s College London 4
D
E
LiklihoodSampling Pla) • Ma bl a Metal . I • I
•W
Plan! ✗
F
D
0.12
P (A) .cat/–Sw–PCd)Ple1rb,#
A P(B|A) T 0.4
0.9
@B D P ( E | B , D )
@TT 0.4 TF 0.5 FT 0.2 FF 0.1
Calculate P pa|e, dq
to o
W=0
6.0.2
.
A
A P(C|A) T 0.5
0.9 BFC
P(D) I 0.6
0.4
E
0.29
DE
PEAOC
random numbers: 0.2, 0.3, 0.4, 0.1, 0.5, 0.2, 0.5, 0.1, 0.3, 0.4, 0.9, 0.9, 0 , 0.1, 0.2 ab cd•_E←-aabc Toe
c -Trenn, King’s College London 0 .1 2 5
W
=
–
=
= 0 6.0.4
Liklihood Sampling
P (A)
0.4
A
P(D)
0.6
A
P(C|A)
T F
0.5 0.9
C
D
A
T F
BD
TT TF FT FF
P(B|A)
0.4 0.9
P(E|B,D)
0.4 0.5 0.2 0.1
B
E
CalculatePpa|e,dq =£
random numbers: 0.2, 0.3, 0.4, 0.1, 0.5, 0.2, 0.5, 0.1, 0.3, 0.4, 0.9, 0.9, 0 , 0.1, 0.2
pcnale, d) =μ←→
c -Trenn, King’s College London 5