As MCCV runs, it incrementally builds a fitness table. The fitness table has one entry for every internal run. The total number of runs it the number of trials times the number of parameters values. Each entry of the fitness table has three elements. The first is the value of the chosen parameters name for that trial, the second value if the number of clusters returned by the clustering algorithm, the third parameters is the fitness as computed by the fitness() method of the Model object internal to the chosen clustering algorithm. An example fictitious fitness table is show in Table 1.
|