proDSinit function

Initialization of parameters for the evidential neural network classifier

proDSinit returns initial parameter values for the evidential neural network classifier.

proDSinit(x, y, nproto, nprotoPerClass = FALSE, crisp = FALSE)

Arguments

• x: Input matrix of size n x d, where n is the number of objects and d the number of attributes.
• y: Vector of class labels (of length n). May be a factor, or a vector of integers from 1 to M (number of classes).
• nproto: Number of prototypes.
• nprotoPerClass: Boolean. If TRUE, there are nproto prototypes per class. If FALSE (default), the total number of prototypes is equal to nproto.
• crisp: Boolean. If TRUE, the prototypes have full membership to only one class. (Available only if nprotoPerClass=TRUE).

Returns

A list with four elements containing the initialized network parameters

• alpha: Vector of length r, where r is the number of prototypes.
• gamma: Vector of length r
• beta: Matrix of size (r,M), where M is the number of classes.
• W: Matrix of size (r,d), containing the prototype coordinates.

Details

The prototypes are initialized by the k-means algorithms. The initial membership values $u_{ik}$ of each prototype $p_i$ to class $\omega_k$ are normally defined as the proportion of training samples from class $\omega_k$ in the neighborhood of prototype $p_i$. If arguments crisp and nprotoPerClass are set to TRUE, the prototypes are assigned to one and only one class.

Examples

## Glass dataset
data(glass)
xapp<-glass$x[1:89,]
yapp<-glass$y[1:89]
param0<-proDSinit(xapp,yapp,nproto=7)
param0

References

T. Denoeux. A neural network classifier based on Dempster-Shafer theory. IEEE Trans. on Systems, Man and Cybernetics A, 30(2):131--150, 2000.

See Also

proDSfit, proDSval

Author(s)

Thierry Denoeux.