You are on page 1of 1

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Scripts %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

%%%%%%%%%%%%%%%%

%% data_prep.m

This file is used for data preprocessing and data preparation for training and
testing

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

%% Main_training.m

This file is the main training file that:


- trains the two HMMs using the training data created in data_prep.m
- model-fitting accuracy of the trained models (to observe over-fitting)
- 10-folds cross-validation accuracy (to observe robustness)

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

%% Main_test.m

This file is used to test the models on the unseen data sequences and predict their
labels

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

%% test_demo.m

This file is used to show the prediction result on a test sequence graphically
- window turns blue for non-hypotensive prediction
- window turns red for hypotensive prediction

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% Functions %%%%%%%%%%%%%%%%%%%%%%%%%%%%%


%%%%%%%%%%%%%%%

%% ourtrainhmm.m

This function estimates the model parameters using Baum-Welch algorithm and
estimates emission prob
using kmeans clustering

- inputs: training data

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

%% ourtesthmm.m

This measures likelihood for a sequence given model parameters

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%%%%%%%%%%%%%%%%%%%%%%

You might also like