Table of Contents

General stuff

• Baum-Welch-Algorithm
• Discrete HMM model
• Forward-Backward-Algorithm
• GHMM Constants
• GHMM-Globals
• HMM-Modell
• Help functions concerning memory and small calculations.
• Help functions for printing.
• Help functions for random values
• Matrix
• Viterbi-Algorithm
• cluster
• model
• rng Initialization for the random number generator
• root finder
• scanner Structure and methods; the functions handle memory and
• scluster
• some calculations concerning gaussian tail function

Functions

• discrime_compute_performance Returns the value of the in this discriminative training algorithm optimised function for a tupel of HMMs and sequencesets.
• discriminative Trains two or more models to opimise the discrimination between the classes in the trainingset.
• ghmm_rng_init
• ghmm_rng_timeseed
• gradescent_compute_expectations computes matrices of n and m variables (expected values for how often a certain parameter from A or B is used) computes Baum-Welch variables implicit
• gradient_descent Trains the model with a set of annotated sequences till convergence using gradient descent.
• hlist_propFwd Propagates list of hypotheses forward by extending each old hypothesis to the possible new hypotheses depending on the states in which the old hypothesis could end and the reachable labels
• kbest Calculates the most probable labeling for the given sequence in the given model using k-best decoding.
• logGammaSum Calculates the logarithm of sum(exp(log_a[j,a_pos])+exp(log_gamma[j,g_pos])) which corresponds to the logarithm of the sum of a[j,a_pos]*gamma[j,g_pos]
• logSum Calculates the logarithm of the sum of the probabilities whose logarithms are stored in the given array
• mes
• mes_ability
• mes_calloc
• mes_copy
• mes_err
• mes_exit
• mes_fflush
• mes_fformat
• mes_fgetc
• mes_fopen
• mes_fprintf
• mes_fputc
• mes_fputs
• mes_fread
• mes_fread_quiet
• mes_fseek
• mes_fseek64
• mes_ftell
• mes_fwrite
• mes_get_std_path
• mes_init
• mes_init_args
• mes_init_logfile
• mes_init_winfct
• mes_insert
• mes_malloc
• mes_move
• mes_realloc
• mes_remove
• mes_rename
• mes_smart
• mes_time
• mes_tmpfile
• mes_va
• mes_win_ability
• mprintf_dyn
• mprintf_va
• mprintf_va_dyn
• pair functions dealing with the emission indices *
• reestimate_alloc_matvek matrix allocation and free for training algorithms

Macros

• mes_check_0
• mes_check_bnd
• mes_check_expr
• mes_check_neg
• mes_check_ptr
• mes_file
• mes_file_win
• mes_proc
• mes_prot
• mes_win

Typedefs

• hypo_List Data type for linked list of hypotheses Stores the actual label, a link to the parent hypothesis, a counter of the links to this hypothesis and the gamma values