Feedback on first draft

Overall this is very good.

What is the V-measure? I presume higher is better. How is it calibrated (e.g., what is the V-measure of random guessing)? Is 55% good or just close to random guessing? (Only put a formal definition in if it improves readability/understandability)

The CopyMix graphical model was difficult to understand. I presume Y1..Y_n are related to Y. Are they really independent or independent given the clustering and the (latent) state of the Markov chain? What is "time" in the Markov chain? Perhaps drawing the graphical model would help. (Also is "CopyMix does not require hyperparameter selection and is not limited by dataset size" really true?)

"Contributions of Ginkgo to CopyMix" doesn't reflect the content of the section (at least the first paragraph). I think this paragraph is trying to answer: what part of Ginko did CopyMix build on? What advances were there, both beyond the techniques used and beyond the techniques (e.g., cluster size)? How is the performance compared?