An implementation of the umbrella world Hidden Markov Model found in the Russel & Norvig book, with direct sampling and the forward-backward algorithm.
NumPy needs to be installed. On any system with pip
installed:
pip install numpy
The sampler.py
script generates a number of sample files, each of which contain a sequence of a certain length (meaning: the events are sampled for some consequent days). You can pass one of those files to model.py
, which takes it as an input and uses the forward-backward algorithm to calculate the posterior marginals based on the observations only. Marginals are stored in marginals.txt
. An accuracy.txt
file is then generated comparing the guesses of the algorithm to the actual event. A "wrong" value doesn't actually mean the marginals are wrong, since it only takes into account whether the algorithm considered one outcome more likely. It only means that the outcome which the algorithm considered less likely happened. If on many tests with different samples there are many more "right" values than "wrong" values, the algorithm is behaving correctly.
Copyright (C) 2018 jotaro-sama
This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this program. If not, see http://www.gnu.org/licenses/.