ng.m

竞争学习的matlab工具箱

% Basic implementation of the Neural-Gas algorithm
%
% Source: 
%    T. M. Martinetz, S. G. Berkovich, and K. J. Schulten.
%    Neural-gas network for vector quantization and its application to time-series prediction.
%    IEEE Transactions on Neural Networks, 4(4):558-569, 1993.
%
% This implementation aims to be simple and direct. More powerful
% implementations of the Neural-Gas can be found in the SOMtoolbox.
% 
% Authors: Guilherme A. Barreto
% Date: November 17th 2005

clear; clc; close all;

% Load data
load dataset1.dat;
Dw=dataset1; clear dataset1

% Get size of data matrix (1 input vector per row)
[LEN_DATA DIM_INPUT]=size(Dw);

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Define size of the network  %%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Mx = 4;   % Number of neurons in the X-dimension
My = 4;   % Number of neurons in the Y-dimension
MAP_SIZE = [Mx My];        % Size of SOM map (always use 1-D map)

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Create a CL network structure  %%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
sMap = som_map_struct(DIM_INPUT,'msize',MAP_SIZE,'rect','sheet');

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Different weights initialization methods %%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% sMap  = som_randinit(Dw, sMap);   % Random weight initialization
% sMap  = som_lininit(Dw, sMap);    % Linear weight initialization
I=randperm(LEN_DATA); sMap.codebook=Dw(I(1:Mx*My),:);  % Select Mx*My data vectors at random

Co=som_unit_coords(sMap); % Coordinates of neurons in the map

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Specification of some training parameters %%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
li=10; %round(max(Mx,My)/2);     % Initial neighborhood
lf=0.01;                % Final neighborhood
ei=0.1;                  % Initial learning rate
ef=0.001;                % Final learning rate
Nep=100;                % Number of epochs
Tmax=LEN_DATA*Nep;     % Maximum number of iterations
T=0:Tmax;              % Time index for training iteration
lambda=li*power(lf/li,T/Tmax);  % Learning rate vector
eta=ei*power(ef/ei,T/Tmax);  % Neighborhood width vector

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
% Train Kohonen Map (TKM)  %%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
for t=1:Nep,  % loop for the epochs
    
    epoch=t, % Show current epoch
    for tt=1:LEN_DATA,
         % Compute distances of all prototype vectors to current input
         Di=sqrt(som_eucdist2(sMap,Dw(tt,:)));

         % Sort Di in ascending order 
         [Di_ordered RANK] = sort(Di);
     
         % Update the weights of the winner and its neighbors
         T=(t-1)*LEN_DATA+tt;    % iteration throughout the epochs
         for i=1:Mx*My,
             % Find the position of neuron "i" in RANK
             ki=find(RANK==i);
             
             % Compute the corresponding weighting function
             H=exp(-(ki-1)/lambda(T));
             
             % Update the weights of neuron "i"
             sMap.codebook(i,:)=sMap.codebook(i,:) + eta(T)*H*(Dw(tt,:)-sMap.codebook(i,:));
         end
    end
    % Quantization error per training epoch
    Qerr(t) = som_quality(sMap, Dw);
end


% Plot prototypes and data altogether
figure, plot(Dw(:,1),Dw(:,2),'+r'), hold on
plot(sMap.codebook(:,1),sMap.codebook(:,2),'b*')
title('Prototype vectors in input space'), hold off

% Plot quantization error evolution per training epoch
figure, plot(Qerr) 
title('Quantization Error per Training Epoch')