Markov Models 

Sad story :(

Yes, in this case you do. Though there are some transition matrices that will not converge when raised to a large power.

She saw this, she doesn't mind :)

Glad you found the video helpful. Those are transition probabilities, and one way to obtain them is to observe the system (Rommie) for a long period of time as she stochastically transitions from a given state (her house) to the other states. Let's say, for example, you observed 20 transitions, and she went to work 10 out of 20 times, you get a 0.5 probability. If she went to her house 8 out of 20 times, you get a 0.4 probability, etc.

@@lanevotapka4012 Hey Lane, thank you so much for your answer.

The starting probability vector is something that you need to construct yourself or is provided to you. Each place in the vector represents the state, and you need to put the probability that the system starts in those states. That means, if we know that Rommie (in the movie) always starts from home, then we put a one in that spot of the starting probability vector, and put zeros everywhere else.

Good question Vijay .

That's why you just let the computer do it for you :)