matlab-基于matlab的算术优化算法(AOA)仿真,通过AOA优化KNN聚类性能,输出AOA优化收敛曲线以及优化后的KN

function AOA = jArithmeticOptimizationAlgorithm(feat,label,opts) % Parameters lb = 0; ub = 1; thres = 0.5; MOP_Max = 1; MOP_Min = 0.2; alpha = 5; Mu = 0.499; if isfield(opts,'N'), N = opts.N; end if isfield(opts,'T'), max_Iter = opts.T; end if isfield(opts,'MOP_Max'), MOP_Max = opts.MOP_Max; end if isfield(opts,'MOP_Min'), MOP_Min = opts.MOP_Min; end if isfield(opts,'alpha'), alpha = opts.alpha; end if isfield(opts,'Mu'), Mu = opts.Mu; end if isfield(opts,'thres'), thres = opts.thres; end %objective function fun = @jFitnessFunction; %Number of dimensions dim = size(feat,2); % Initial X = zeros(N,dim); % Here I'am having small doubt for i = 1:N for d = 1:dim X(i,d) = lb + (ub - lb) * rand(); end end %Fitness fit = zeros(1,N); fitG = inf; %Calculate the fitness values of solutions for i=1:N fit(i)=fun(feat,label,(X(i,:)>thres),opts); %having doubt here if fit(i)<fitG fitG=fit(i); Xgb=X(i,:); end end Xnew = zeros(N,dim); %Fnew = zeros(1,size(Xnew,1)); % havind doubt in this particular line % curve = zeros(1,max_Iter); t=1; while t<max_Iter+1 %Main loop MOP=1-((t)^(1/alpha)/(max_Iter)^(1/alpha)); % Probability Ratio MOA=MOP_Min+t*((MOP_Max-MOP_Min)/max_Iter); %Accelerated function %Update the Position of solutions for i=1:N % if each of the UB and LB has a just value for j=1:dim r1=rand(); if (size(lb,2)==1) if r1<MOA r2=rand(); if r2>0.5 Xnew(i,j)=Xgb(1,j)/(MOP+eps)*((ub-lb)*Mu+lb); else Xnew(i,j)=Xgb(1,j)*MOP*((ub-lb)*Mu+lb); end else r3=rand(); if r3>0.5 Xnew(i,j)=Xgb(1,j)-MOP*((ub-lb)*Mu+lb); else Xnew(i,j)=Xgb(1,j)+MOP*((ub-lb)*Mu+lb); end end end if (size(lb,2)~=1) % if each of the UB and LB has more than one value r1=rand(); if r1<MOA r2=rand(); if r2>0.5 Xnew(i,j)=Xgb(1,j)/(MOP+eps)*((ub(j)-lb(j))*Mu+lb(j)); else Xnew(i,j)=Xgb(1,j)*MOP*((ub(j)-lb(j))*Mu+lb(j)); end else r3=rand(); if r3>0.5 Xnew(i,j)=Xgb(1,j)-MOP*((ub(j)-lb(j))*Mu+lb(j)); else Xnew(i,j)=Xgb(1,j)+MOP*((ub(j)-lb(j))*Mu+lb(j)); end end end end Flag_UB=Xnew(i,:)>ub; % check if they exceed (up) the boundaries Flag_LB=Xnew(i,:)<lb; % check if they exceed (down) the boundaries Xnew(i,:)=(Xnew(i,:).*(~(Flag_UB+Flag_LB)))+ub.*Flag_UB+lb.*Flag_LB; Fnew =fun(feat,label,(Xnew(i,:)),opts); % calculate Fitness function if Fnew <fit(i) X(i,:)=Xnew(i,:); fit(i)=Fnew; end if fit(i)<fitG fitG=fit(i); Xgb=X(i,:); end end %Update the convergence curve curve(t)=fitG; fprintf('\nIteration %d Best (AOA)= %f',t,curve(t)) t=t+1; % incremental iteration end % select features based on selected index Pos = 1:dim; Sf = Pos((Xgb > thres) == 1); sFeat = feat(:,Sf); % Store results AOA.sf = Sf; AOA.ff = sFeat; AOA.nf = length(Sf); AOA.c = curve; AOA.f = feat; AOA.l = label; end