Calculates the
Quantisation Error, defined as the average distance between and input data vector and the
weight-vector of its best-matching-unit.
Calculates the following values:
- Unit
- Quantisation error (qe). Calculated for each unit as the sum of the distances between the input vectors mapped to
this unit to the unit's weight vector.
- Mean quantisation error (mqe). Calculated for each unit as the (qe/|mapped vectors|), i.e. the qe divided by the
number of input data vectors mapped on this unit.
- Map
- Mean quantisation error (mqe), caluclated as the sum of all unit qe's divided by the number of units with at
least one mapped input data vector.
- Mean mean quantisation error (mmqe), calculated as the sum of all unit mqe's divided by the number of units with
at least one mapped input data vector.