Demo & applications

A simple demo is provided here. A set of one-dimensional signals generated from a HMM with Gaussian distributions is used to learn back the HMM parameters using a maximum likelihood training.

In order to run the demo go to the folder demo, open a julia REPL and type the following:

julia> using Pkg; Pkg.activate("."); Pkg.instantiate()

julia> include("demo.jl")

Other applications:

• Automatic speech recognition: TIDIGITS recipe