markov | Text generation using Markov Chains | Development Tools library

This is my Python script for using Markov Blanket Algorithm on my Dataset:

I have a multivariate model with this (approximate) form:

I have the following kind of data:

I wrote a little R snippet to go over a vector containing realisations from a Markov chain and return the observed transitions for a given order. For concreteness, suppose we're interested in order 2 transitions for a state-space $\mathcal{S}$. The ultimate goal is to store the counts $n_{ijk}$, $i, j, k \in \mathcal{S}$ in a convenient form for later use.

Suppose I have the following numpy array:

I am struggling to find an answer to this question,i have not understand well Markov chain use and i would like some help:

[Input image] https://i.stack.imgur.com/r9XCE.png

[Output image example] https://i.stack.imgur.com/3pllU.png

This is what we begin with, classic.

I want to transform this image in black and white only BUT i have to use Markov chain probabilities which I don't understand at all

Is anyone able to give me somme hint? I believe that we have to pick random pixel N number of time and to apply somme proba magic to change it to black or white (based on probabilities and neighbors)

thanks in advance (i'm not searching for code particulary but the logic to be implemented)

Using the great TensorFlow Hidden Markov Model library, it is straightforward to model the following Dynamic Bayesian Network:

where H_i is the probability variable that represents the HMM and S_i is the probability variable that represents observations.

What if I'd like to make H depend on yet another HMM (Hierarchical HMM) or simply other probability variable like this:

The HiddenMarkovModel definition in TensorFlow looks like the following:

I am trying to simulate a step through a Markov chain with the intent to loop through the procedure multiple times until a condition is met. (i.e., find out how many steps it takes, on average, to reach a specific state).

In this case, a state can only go one way. E.g., State 4 can transition forward to State 5, but cannot transition backward to State 3. This means the left-lower half of the transition matrix is set to zero. This is also why the method below puts arbitrarily large values in the 'prior' states. I attempt to find the correct new state by examining which probability in the specified row of the transition matrix is closest to a random number.

I have a data frame with a column named title, I want to apply textdistance to check similarities between different titles and remove any rows with similar titles (based on a specific threshold). Is there away to do that directly, or I need to define a custom function and group similar titles togother before removing "duplicates" (titles that are similar)? A sample would look like this.