In the Forward-backward algorithm, this Transition probabilities:

$\mathbf {T} ={\begin{pmatrix}0.7&0.3\\0.3&0.7\end{pmatrix}}$

could be viewed as the probabilities of the weather at t+1, for instance, $T_{11}$ could be viewed as

the weather has a 70% chance of staying the same

the emission probabilities

$\mathbf {B} ={\begin{pmatrix}0.9&0.1\\0.2&0.8\end{pmatrix}}$

could also be interpreted as

90% people use umbrella when it is rainy

in this context, what does following observations mean?

$\mathbf {O_{1}} ={\begin{pmatrix}0.9&0.0\\0.0&0.2\end{pmatrix}}~~\mathbf {O_{2}} ={\begin{pmatrix}0.9&0.0\\0.0&0.2\end{pmatrix}}~~\mathbf {O_{3}} ={\begin{pmatrix}0.1&0.0\\0.0&0.8\end{pmatrix}} ...$

  • $\begingroup$ Don't forget to include links or references, so people know what you are talking about. $\endgroup$ – Graham Kemp Jul 25 '19 at 3:41

To be precise, $\mathbf T_{11}$ is the conditional probability that: the weather stays in state 1 (when it is in state 1).   It is that we have $\mathbf T_{22}=\mathbf T_{11}$ which indicates that the event that the weather stays the same is independent of the current state of the weather.

The diagonal elements of $\mathbf O_j$ present the conditional probabilities for observing event $j$ given each system state. These events are given in the array, $\{$umbrella, umbrella, no-umbrella, umbrella, umbrella$\}$.

So $\mathbf O_1$ presents us with the conditional probability for observing an umbrella would be 0.9 when given a rainy day, and 0.2 when given a sunny day.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.