# 34 this is just a dierent semi ring the resulting

This preview shows page 1. Sign up to view the full content.

This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: observations does not in general factorize, so no i.i.d. here either. The distribution pk (x1:T ) also does not in general have any vit (marginal) independence assumptions. Prof. Jeﬀ Bilmes EE596A/Winter 2013/DGMs – Lecture 5 - Jan 25th, 2013 page 5-55 (of 232) HMMs Trellis Other HMM queries MPE Sampling What HMMs can do Summary Scratch HMMs and stationarity HMM is a stochastic process over {Xt }t Prof. Jeﬀ Bilmes EE596A/Winter 2013/DGMs – Lecture 5 - Jan 25th, 2013 page 5-56 (of 232) HMMs Trellis Other HMM queries MPE Sampling What HMMs can do Summary Scratch HMMs and stationarity HMM is a stochastic process over {Xt }t An HMM is stationary whenever (5.35) p(Xt1+h:n+h = x1:n ) = p(Xt1:n = x1:n ) When might this be the case? Prof. Jeﬀ Bilmes EE596A/Winter 2013/DGMs – Lecture 5 - Jan 25th, 2013 page 5-56 (of 232) HMMs Trellis Other HMM queries MPE Sampling What HMMs can do Summary Scratch HMMs and stationarity HMM is a stochastic process over {Xt }t An HMM is stationary whenever (5.35) p(Xt1+h:n+h = x1:n ) = p(Xt1:n = x1:n ) When might this be the case? We have p(Xt1:n +h = x1:n ) Prof. Jeﬀ Bilmes EE596A/Winter 2013/DGMs – Lecture 5 - Jan 25th, 2013 page 5-56 (of 232) HMMs Trellis Other HMM queries MPE Sampling What HMMs can do Summary Scratch HMMs and stationarity HMM is a stochastic process over {Xt }t An HMM is stationary whenever (5.35) p(Xt1+h:n+h = x1:n ) = p(Xt1:n = x1:n ) When might this be the case? We have p(Xt1:n +h = x1:n ) p(Xt1:n +h = x1:n , Qt1:n +h = q1:n ) = q1:n Prof. Jeﬀ Bilmes EE596A/Winter 2013/DGMs – Lecture 5 - Jan 25th, 2013 page 5-56 (of 232) HMMs Trellis Other HMM queries MPE Sampling What HMMs can do Summary Scratch HMMs and stationarity HMM is a stochastic process over {Xt }t An HMM is stationary whenever (5.35) p(Xt1+h:n+h = x1:n ) = p(Xt1:n = x1:n ) When might this be the case? We have p(Xt1:n +h = x1:n ) p(Xt1:n +h = x1:n , Qt1:n +h = q1:n ) = q1:n = p(Qt1 +h = q1 )p(Xt1 +h = x1 |Qt1 +h = q1 ) q1:n n i=2 Prof. Jeﬀ Bilmes p(Xti +h = xi |Qti +h = qi )p(Qti +h = qi |Qti−1 +h = qi−1 ) EE596A/Winter 2013/DGMs – Lecture 5 - Jan 25th, 2013 page 5-56 (of 232) HMMs Trellis Other HMM queries MPE Sampling What HMMs can do Summary Scratch HMMs and stationarity (cont.) . . . continuing = q1 p(Qt1 +h = q1 )p(Xt1 +h = x1 |Qt1 +h = q1 ) (5.36) n q2:T i=2 p(Xti +h = xi |Qti +h = qi )p(Qti +h = qi |Qti−1 +h = qi−1 ) (5.38) Prof. Jeﬀ Bilmes EE596A/Winter 2013/DGMs – Lecture 5 - Jan 25th, 2013 page 5-57 (of 232) HMMs Trellis Other HMM queries MPE Sampling What HMMs can do Summary Scratch HMMs and stationarity (cont.) . . . continuing = q1 p(Qt1 +h = q1 )p(Xt1 +h = x1 |Qt1 +h = q1 ) (5.36) n q2:T i=2 = q1 p(Xti +h = xi |Qti +h = qi )p(Qti +h = qi |Qti−1 +h = qi−1 ) p(Qt1 +h = q1 )p(Xt1 +h = x1 |Qt1 +h = q1 ) (5.37) n q2:T i=2 p(Xti = xi |Qti = qi )p(Qti = qi |Qti−1 = qi−1 ) (5.38) Prof. Jeﬀ Bilmes EE596A/Winter 2013/DGMs – Lecture 5 - Jan 25th, 2013 page 5-57 (of 232) HMMs Trellis Other HMM queries...
View Full Document

Ask a homework question - tutors are online