HMM Basic problem 3 Given labelled data sequence D=X,, a1, Z1,X2, a2, Z2 XT, aT, ZT3 estimating p(z, Xi) and p(xi X, a) is just counting Given unlabelled data sequence D={a1,z1a2,z2,…,a,Z1} estimating p(Zi xi) and p(xi X; ak is equivalent to simultaneous localization and mapping (next lecture) Particle FiltersHMM Basic Problem 3 ● D={x1,a1,z1,x2,a2,z2,...,xT,aT,zT}, estimating p(zi ,xi ) and p(xj |xi ,ak) is just counting ● Given unlabelled data sequence, D={a1,z1,a2, z2, ...,aT,zT}, estimating p(zi ,xi ) and p(xj |xi ,ak) is equivalent to simultaneous localization and mapping (next lecture) Particle Filters Given labelled data sequence