What is Hidden Markov models (HMMs)?

A Hidden Markov Model is like having a friend who tells you what they're doing, but not why, and you have to guess what's going on behind the scenes.

Imagine your friend is playing with toys in another room. You can only see what they tell you: "I’m playing with blocks!" or "I’m drawing!". But you don’t know if they’re happy, tired, or just switching toys. That’s like a Hidden Markov Model, you see the result (what your friend is doing), but not the reason (why they chose that activity).

Think of it as a secret toy box with different sections: blocks, crayons, and puzzles. Your friend picks a section at random each time, and then tells you what they’re playing with.

• If they pick blocks, they say "I’m building a tower!"
• If they pick crayons, they say "I’m drawing a dragon!"
• If they pick puzzles, they say "I’m solving a mystery!"

You can’t see the toy box, but you can guess what section your friend picked based on what they tell you. That’s how Hidden Markov Models work, guessing hidden states (like the toy sections) from visible results (what your friend says).