The "Speech Processing Rabiner Solution" is not a PDF. It is the discipline of quantifying the uncertainty of a speech signal . It is the realization that speech is a time-varying process, and any mathematical model is just an approximation.
Now you have a sequence of feature vectors. How do you match them to words? Speech Processing Rabiner Solution