Announcements
● Final Wed May 11th 6pm, location: TBD
● Lab 3: due Thursday
Hate(Jim, Dog(Fido)) # illegal Hate(Jim, Fido) ^ Dog(Fido) Hate(Sk1(x), x)
Copyright By PowCoder代写 加微信 powcoder
For-all x, there-exists y Hate(y, x)
P(C=T|P=T, Q=T) = P(C=T) * P(P=T|C=T) * P(Q=T|C=T) = .6 *2/30 * 1⁄2 = .02 P(C=F|P=T, Q=T) = P(C=F) * P(P=T|C=F) * P(Q=T|C=F) = 2/5 * 2/5* 1⁄4 = .04
P(C=T) = 30/50 P(C=F) = 20/50 P(P=T|C=T) = 2 / 30 P(Q=T|C=T) = 15/30 P(P=T|C=F) = 8/20 P(Q=T|C=F) = 5/20
log (P(C=F|P=T, Q=T)) = log(P(C=F)) + log(P(P=T|C=F)) + log( P(Q=T|C=F))
def ValueIteration(P): V = all zeros
for num iterations:
V’ = R + df* P * V
if (for each v, v’ in (V, V’) abs(v-v’) < tol ) return V’ V = V’
def prog():
P = some arbitrary P While true:
V = ValueIteration(P) P’ = recompute using V If P == P’ return P, V
K : [G, J, L, P]
K = -1 + .9 *( .85 * P + .05 *G + .05* L + .05 * J) K=-1+.9*(.8*Policy+.05*(G+L+P+ J)) Argmax [0, 1, 5, 3] => i=2
Policy(K) = argmax (G, J, L, P) -> {K -> J} argmax(G=0.2, J=0.1, L=0.5, P=0)-
If ‘-min’ flag, then use argmin
Tol = 0.001 yours to be +- 0.01
程序代写 CS代考 加微信: powcoder QQ: 1823890830 Email: powcoder@163.com