👨‍🎓个人主页：研学社的博客 

💥💥💞💞欢迎来到本博客❤️❤️💥💥

🏆博主优势：🌞🌞🌞博客内容尽量做到思维缜密，逻辑清晰，为了方便读者。

⛳️座右铭：行百里者，半于九十。

📋📋📋本文目录如下：🎁🎁🎁

💥1 概述

特征选择（特征子集选择）问题是各个领域重要的预处理阶段之一。在真实的数据集中，存在许多不相关、误导性和冗余的特征，这些特征是无用的。主要特征可以通过特征选择技术提取。特征选择属于 NP 难题类;因此，元启发式算法对于解决问题很有用。引入一种新的二元ABC称为二元多邻域人工蜂群（BMNABC），以增强ABC阶段的勘探和开发能力。BMNABC 在第一阶段和第二阶段使用新的概率函数应用近邻域和远邻域信息。在第三阶段，对那些在前一阶段没有改进的解决方案进行了比标准ABC更有针对性的搜索。该算法可以与包装器方法结合使用，以获得最佳结果。

📚2 运行结果

 部分代码：

function [BestChart,Error,BestFitness,BestPosition]= BMNABC(FoodNumber,maxCycle,D,fobj,Limit)
global SzW
rand('state',sum(100*clock)) %#ok
%reset trial counters
trial=zeros(1,FoodNumber);
rmin=0;rmax=1;
Vmax=6;
BestChart=[];

%%
%/FoodNumber:The number of food sources equals the half of the colony size*/
Foods=initialization(FoodNumber,D,1,0)>0.5;

for i=1:FoodNumber
    ObjVal(i) =feval(fobj,Foods(i,:));
end

BestInd=find(ObjVal==min(ObjVal));
BestInd=BestInd(end);
BestFitness=ObjVal(BestInd);
BestPosition=Foods(BestInd,:);
x=BestPosition;
Acc=(SzW*sum(x)/length(x)-BestFitness)/(1-SzW)+1; % This relation is computed based on FitnessValue=(1-SzW)*(1-cp.CorrectRate)+SzW*sum(x)/(length(x));
% The feature subset is based on two criteria: the maximum classi?cation accuracy (minimum error rate) and the minimal number of selected features. 
%Fitness = a* ClassificationError+ (1-a) NumberOfSelectedFeatures/TotalFeatures

 
Error=1-Acc;
BestChart=[BestChart,Error];
fprintf('Iteration:%3d\tAccuracy:%7.4f\tFitness:%7.4f\t#Selected Features:%d\n',1,((1-Error)*100),BestFitness,(sum(BestPosition,2)));

gindex=BestInd;

pbest=Foods;
pbestval=ObjVal;
iter=2;
while ((iter <=maxCycle)),   
    r= rmax-iter*(rmax-rmin)/maxCycle;
    
    %%
    %%%%%%%%% EMPLOYED BEE PHASE %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
    for i=1:(FoodNumber)
        neigh=[]; d=[]; dd=[];
        w=0;
        for k=1:FoodNumber
            if k~=i
                w=w+1;
                d(w)=sum(Foods(i,:)~=Foods(k,:)); %Hamming distance.
                dd(w)=k;
            end
        end
        Meand=mean(d);
        Maxd=max(d);
        u=1;
        for z=1:w
            
            if (d(z)<=Maxd && d(z)>=r*Meand)
                
                neigh(u)=dd(z);
                u=u+1;
            end
        end
        
        
        
        fi=rand(length(neigh),D);
        
        FIP=sum(fi.*pbest(neigh,:))./sum(length(neigh));
        sol=Foods(i,:);
        
        V(i,:)=rand(1,D).*(Foods(i,:)-FIP);
        V(i,:)=max(min(V(i,:),Vmax),-Vmax);
        aa=abs(tanh(V(i,:)));
        
        AA=1-exp(-trial(i)/(iter*maxCycle));
        S=AA+aa;
        
        temp=rand(1,D)         moving=find(temp==1);
        
        sol(moving)=~sol(moving);
        
        
        ObjValSol=feval(fobj,sol);
        
        if (ObjValSol             pbest(i,:)=sol;
            pbestval(i)=ObjValSol;
        end
        
        if (ObjValSol             Foods(i,:)=sol;
            ObjVal(i)=ObjValSol;
            trial(i)=0;
        else
            trial(i)=trial(i)+1; %/*if the solution i can not be improved, increase its trial counter*/
        end;
        
        
    end;
    
    %%
    %%%%%%%%%%%%%%%%%%%%%%%% ONLOOKER BEE PHASE %%%%%%%%%%%%%%%%%%%%%%%%%%%
    
    i=0;
    
    while(i         i=i+1;
        
        neigh=[]; d=[];dd=[];
        w=0;
        for k=1:FoodNumber
            if k~=i
                w=w+1;
                d(w)=sum(Foods(i,:)~=Foods(k,:));
                dd(w)=k;
            end
        end
        Meand=mean(d);
        u=1;
        for z=1:w
            
            if (d(z)<=Meand)
                neigh(u)=dd(z);
                u=u+1;
            end
        end
        
        
        [tmp,tmpid]=min(ObjVal(neigh));
        
        nn=neigh(tmpid);
        
        sol=Foods(i,:);
        
        
        V(i,:)=rand(1,D).*(Foods(i,:)-Foods(nn,:));
        V(i,:)=max(min(V(i,:),Vmax),-Vmax);
        aa=abs(tanh(V(i,:)));
        
        
        AA=1-exp(-trial(i)/(iter*maxCycle));
        S=AA+aa;
        
        temp=rand(1,D)         moving=find(temp==1);
        
        sol(moving)=~sol(moving);
        
        
        ObjValSol=feval(fobj,sol);
        
        if (ObjValSol             pbest(i,:)=sol;
            pbestval(i)=ObjValSol;
        end
        % /*a greedy selection is applied between the current solution i and its mutant*/
        if (ObjValSol             Foods(i,:)=sol;
            ObjVal(i)=ObjValSol;
            trial(i)=0;
        else
            trial(i)=trial(i)+1; %/*if the solution i can not be improved, increase its trial counter*/
        end;
        
        
    end
   %% 
    %/*The best food source is memorized*/
    
    ind=find(ObjVal==min(ObjVal));
    ind=ind(end);
    if (ObjVal(ind)         
        BestFitness=ObjVal(ind);
        BestPosition=Foods(ind,:);
        x=BestPosition;
        Acc=(SzW*sum(x)/length(x)-BestFitness)/(1-SzW)+1;
        Error=1-Acc;
        gindex=ind;
        success=1;
    else
        success=0;
    end;
    
   %% 
    %%%%%%%%%%%% SCOUT BEE PHASE %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
    
    %/*determine the food sources whose trial counter exceeds the "Limit" value.
    %In Basic ABC, only one scout is allowed to occur in each cycle*/
    
    ind=find(trial==max(trial));
    ind=ind(end);
    if (trial(ind)>Limit)
        
        sol=Foods(ind,:);
        b=randi([1 FoodNumber]);
        while ((b==ind)|| (b==gindex))
            b=randi([1 FoodNumber]);
        end
        
        
        if (success==1)
            sol=or(Foods(ind,:),BestPosition) ;
        else
            sol= or (xor(Foods(ind,:),BestPosition) ,or(Foods(ind,:),Foods(b,:)));
        end
        
        ObjValSol=feval(fobj,sol);
        
        if (ObjValSol             pbest(ind,:)=sol;
            pbestval(ind)=ObjValSol;
        end
        if (ObjValSol             Foods(ind,:)=sol;
            ObjVal(ind)=ObjValSol;
            trial(ind)=0;
        else
            trial(ind)=trial(ind)+1; %/*if the solution i can not be improved, increase its trial counter*/
        end;
    end
    
    BestChart=[BestChart,Error];
    fprintf('Iteration:%3d\tAccuracy:%7.4f\tFitness:%7.4f\t#Selected Features:%d\n',iter,((1-Error)*100),BestFitness,(sum(BestPosition,2)));
    
    iter=iter+1;
end

🎉3 参考文献

部分理论来源于网络，如有侵权请联系删除。

Zahra Beheshti(2018), BMNABC: Binary Multi-Neighborhood Artificial Bee Colony for High-Dimensional Discrete  Optimization Problems, Cybernetics and Systems 49 (7-8), 452-474.
 

🌈4 Matlab代码实现