Goto:

Fields, Methods, Tagged Values

See also:

Kneuron, DBDataAdmin, ColumnsFilter, Neighbour, Neighbours, NeighbourhoodShape, KohonenNetworkArgs, SOMForm, Umatrix

Appears in:

CD Kohonen network, CD Umatrix, CD Main module

Connections

Association link to class [CD Kohonen network].Kneuron     Association link to class [Class Diagram].DBDataAdmin     Association link to class [CD Kohonen network].ColumnsFilter     Dependency link to class [CD Kohonen network].Neighbour     Dependency link to class [CD Kohonen network].Neighbours     Association link to class [CD Kohonen network].NeighbourhoodShape     Dependency link to class [CD Kohonen network].KohonenNetworkArgs     Association link from class [CD Main module].SOMForm     Association link from class [CD Umatrix].Umatrix

KohonenNetwork Attributes

Attribute Details private int   x Initial: 5 private int   y Initial: 5 private int   patternLength Initial: 0 private int   iterations Initial: 200 private int   trainDataCount private double   K private double   neighbourhood Initial: 3 private double   neighbourhood2 private double   alpha private double   alpha2 private double   lambda private NeighbourhoodShape   neighShape Initial: NeighbourhoodShape.hexagonal public Kneuron   network private DBDataAdmin   dbAdmin Attribute=[NonSerialized] private int   tableVersion Initial: -1 Attribute=[NonSerialized] private int   umatrixID private int   to_do public ColumnsFilter   usedColumns Attribute=[NonSerialized] public WeightChangeHandler   OnWeightChange <<event>> public DrawObjectEventHandler   OnDrawObject <<event>> public DBDataAdmin   DBAdmin Initial: value Notes: access ti database

KohonenNetwork Methods

Operation Details public KohonenNetwork(): Notes: Constructor public KohonenNetwork(    int netX,    int netY,    double alpha,    double K,    double neighbourhoodSize,    NeighbourhoodShape neighbourhoodShape,    int numberOfIteration,    DBDataAdmin dbAdmin,    int tableVersion): Notes: Constructor with some parameters public UsedColumns():ColumnsFilter <<property>> Notes: Info about which columns are selected. public SelectedCorrectColumns():bool <<property>> Notes: true if Int and Double columns are selected, else false public ExistVersion():bool <<property>> Notes: Test if exist version of data public netX():int <<property>> Notes: Size of network in X - axis public netY():int <<property>> Notes: Size of network in Y - axis public TrainDataCount():int <<property>> Notes: Count of train data == count of row in table public Alpha():double <<property>> Notes: Alpha - training parameter public Cooling():double <<property>> Notes: Cooling - training parameter public Neighbourhood():double <<property>> Notes: Size of neighbourhood public PatternLength():int <<property>> Notes: Length of trainign pattern with unused columns public NumberOfIteration():int <<property>> Notes: Count of algorithm iterations public NeighShape():NeighbourhoodShape <<property>> Notes: shape of neighbourhood

KohonenNetwork Tagged Values

Tag Value Notes Attribute [Serializable]

class: KohonenNetwork