使用粒子群优化改进的ElM神经网络分类MATLAB实现代码示例

% function [time, Et]=ANFISELMELM(adaptive_mode) ​ tic; clc; clearvars -except adaptive_mode; close all; adaptive_mode = 'none'; ​ % load dataset data = csvread('iris.csv'); input_data = data(:, 1:end-1); output_data = data(:, end); ​ % Parameter initialization [center,U] = fcm(input_data, 3, [2 100 1e-6]); %center = center cluster, U = membership level [total_examples, total_features] = size(input_data); class = 3; % [Changeable] epoch = 0; epochmax = 400; % [Changeable] %Et = zeros(epochmax, 1); [Yy, Ii] = max(U); % Yy = max value between both membership function, Ii = the class corresponding to the max value ​ % Population initialization pop_size = 10; population = zeros(pop_size, 3, class, total_features); % parameter: population size * 6 * total classes * total features velocity = zeros(pop_size, 3, class, total_features); % velocity matrix of an iteration ​ c1 = 1.2; c2 = 1.2; original_c1 = c1; original_c2 = c2; r1 = 0.4; r2 = 0.6; max_c1c2 = 2; ​ % adaptive c1 c2 % adaptive_mode = 'none'; % class(adaptive_mode) iteration_tolerance = 50; iteration_counter = 0; change_tolerance = 10; is_first_on = 1; is_trapped = 0; %out_success = 0; for particle=1:pop_size   a = zeros(class, total_features);   b = repmat(2, class, total_features);   c = zeros(class, total_features); ​   for k =1:class       for i = 1:total_features % looping for all features           % premise parameter: a           aTemp = (max(input_data(:, i))-min(input_data(:, i)))/(2*sum(Ii' == k)-2);           aLower = aTemp*0.5;           aUpper = aTemp*1.5;           a(k, i) = (aUpper-aLower).*rand()+aLower; ​           %premise parameter: c           dcc = (2.1-1.9).*rand()+1.9;           cLower = center(k,total_features)-dcc/2;           cUpper = center(k,total_features)+dcc/2;           c(k,i) = (cUpper-cLower).*rand()+cLower;       end     % calculate fitness value and update pBest   for i=1:pop_size       particle_position = squeeze(population(i, :, :, :));       e = get_fitness(particle_position, class, input_data, output_data);       if e < pBest_fitness(i)           pBest_fitness(i) = e;           pBest_position(i, :, :, :) = particle_position;       end   end     % find gBest [gBest_fitness, idx] = min(pBest_fitness);   gBest_position = squeeze(pBest_position(idx, :, :, :));     Et(epoch) = gBest_fitness;     % Adaptive c1 and c2   if strcmpi('none', adaptive_mode) == 0       if (epoch > 1) && (Et(epoch) == Et(epoch-1))           iteration_counter = iteration_counter + 1;       else           iteration_counter = 0;           c1 = original_c1;           c2 = original_c2; %             if is_trapped == 1 %                 out_success = out_success + 1; %                 is_trapped = 0; %             end       end ​       if iteration_counter > iteration_tolerance %             is_trapped = 1;           if is_first_on == 1               iteration_left_since_on = (epochmax - epoch) - change_tolerance;               epoch_when_on = epoch;               is_first_on = 0;           end           curr_pos = (epoch-epoch_when_on);           if curr_pos <= iteration_left_since_on               if strcmpi('c1', adaptive_mode) == 1                   c1 = original_c1 + (max_c1c2 - original_c1)/iteration_left_since_on*(curr_pos);                   %fprintf('iter %d - %d\n', epoch, c1)               elseif strcmpi('c2', adaptive_mode) == 1                   c2 = original_c2 + (max_c1c2 - original_c2)/iteration_left_since_on*(curr_pos);                   %fprintf('iter %d - %d\n', epoch, c2)               elseif strcmpi('c1c2', adaptive_mode) == 1                   c1 = original_c1 + (max_c1c2 - original_c1)/iteration_left_since_on*(curr_pos);                   c2 = original_c2 + (max_c1c2 - original_c2)/iteration_left_since_on*(curr_pos);               end           end       end   end     % Draw the SSE plot   plot(1:epoch, Et);   title(['Epoch ' int2str(epoch) ' -> MSE = ' num2str(Et(epoch))]);   grid   pause(0.001); end ​ %[out output out-output] % ---------------------------------------------------------------- time = toc;