Evo eftimov on

Customer reviews
That's right! It is a good idea. Call for active discussion.
I am glad that the blog is constantly evolving. This post just adds popularity.
What does it mean?

Evo eftimov onUsing Hidden Markov Models (HMM) for Prediction

Some of the more widely used (and written about) applications of HMM are for Learning and Inference (Pattern Recognition). This is done by learning the HMM parameters (Initial State Distribution, Sate Transition Matrix, Emission Matrix) from the historical data and then performing computations like determining the current state from the current emissions vector.

But how do you use an HMM for Prediction.

1. Determine the most likely Current State by applying the learned HMM to the data up to the current moment of time.

2. Use the State Transition Matrix to determine the most likely Transition from the Current State to the Next Future State.

3. Use the Emissions Matrix to determine the Emissions Vector (in effect the future / predicted data points) for the Next Future State