Ciao Dendro,
Ecco la mia scansione (100 estrazioni e ruota prese a caso da mia moglie!!), proviamo a fare una previsione e poi vedremo i risultati, premetto che anche io sono all’oscuro di che ruota si tratti……. È mia moglie che essendo scettica vuol mettere alla prova °NNclust°.
Ecco i parametri impostati:
OBSERVATION = 100
NUMBER OF VARIABLES = 5
# NEURONS IN THE MAP = 5
TRAINING CYCLES = 500
RANDOMIZE = NO
LEARNING PARAMETER:
START VALUE = 0.9
END VALUE = 0.1
DECAY = EXPONENTIAL
GAUSSIAN NEIGHBORHOOD:
START VALUE = 50.0%
END VALUE = 1.0%
DECAY = EXPONENTIAL
Ecco il risultato dellelaborazione:
CLUSTER SIZE
CL1 --- CL2 --- CL3 --- CL4
01------ 01------ 10 ------ 88
CLUSTER POSITION ON THE GRID
CL1 --- CL2 --- CL3 --- CL4
ROW ------ 01------ 01------ 01------ 01
COLUMN 01------ 03------ 04------ 05
CLUSTER MEANS
OVE -- CL1 --- CL2 --- CL3 --- CL4
Y1 --- 44.7 --- 13.0 --- 14.0 --- 19.3 --- 48.3
Y2 --- 49.8 --- 03.0 --- 02.0 --- 45.9 --- 51.4
HDR3 --- 45.9 --- 02.0 --- 01.0 --- 16.3 --- 50.3
Y4 --- 43.9 --- 87.0 --- 72.0 --- 69.7 --- 40.2
Y5 --- 42.7 --- 44.0 --- 19.0 --- 32.3 --- 44.1
CLUSTER VARIANCES
OVE --- CL1 --- CL2 --- CL3 --- CL4
Y1 --- 775.4 --- 00.0 --- 00.0 --- 253.1 --- 746.4
Y2 --- 752.2 --- 00.0 --- 00.0 --- 773.4 --- 720.3
HDR3 --- 666.5 --- 00.0 --- 00.0 --- 112.0 --- 581.7
Y4 --- 630.7 --- 00.0 --- 00.0 --- 248.5 --- 570.9
Y5 --- 854.8 --- 00.0 --- 00.0 --- 656.6 --- 884.0
Spero che sia tutto correttto ……..
Saluti
Sincronone
- Pian Piano la Goccia d'Acqua buca la Roccia ! -