www-ai.cs.tu-dortmund.de/LEHRE/VORLESUNGEN/NLS/WS1415/Literatur/Graphical_Models.pptx
PowerPoint-Präsentation
image4.emf
dMRF = kii=1
n ∑⎡⎣⎢
⎤ ⎦⎥+ ki ⋅ kj(i, j )∈E∑⎡⎣⎢
⎤ ⎦⎥
d
MRF
=k
i
i=1
n
å
é
ë
ê
ù
û
ú
+k
i
×k
j
(i,j)ÎE
å
é
ë
ê
ù
û
ú
image5.emf
µ = 1 N
φ(!xi ) i=1
N
∑
oleObject3.bin
oleObject4.bin [...] 179,178,201,200,firefox,/etc/hosts,4096, 361
3,read,1812,179,178,201,200,firefox,/etc/hosts,4096, 0
4,close,1812,179,178,201,200,firefox,/etc/hosts
timestamp, syscall, thread-id, process-id, parent, user [...] calculations for fixed x from 1. for updating the model. K threads.
Optimization l(q)=p(y,x). B threads.
4
Naïve Bayes is almost as accurate as are CRF, much faster, and uses less memory.
Can we do better? Let’s …