Map of parameters to their sufficient statistics. Expectation Maximization determines the parameterMap automatically from the parameters.
1 probability and a vector of zeros for all parameters.
1 probability and a vector of zeros for all parameters. The vector for a parameter must be of length equal to number of possible observations of the parameter.
Probabilities are multiplied using standard multiplication.
Probabilities are multiplied using standard multiplication. Sufficient statistics for each parameter are summed together.
Probabilities are added using standard addition.
Probabilities are added using standard addition. Sufficient statistics for each parameter are weighted by their respective probabilities and summed together, then divided by the sum of both probabilities.
Sum of many entries.
Sum of many entries. Typically, this would be implemented by the ordinary sum, but there may be more efficient implementations.
0 probability and a vector of zeros for all parameters.
0 probability and a vector of zeros for all parameters. The vector for a parameter must be of length equal to number of possible observations of the parameter.
Sum and product operations defined for sufficient statistics. Statistics consist of a probability and counts of the number of times various values have been seen.