1 简介支持向量机 (Support Vector Machines, SVM) 是一种应用广泛的机器学习方法, 具有理论知识清晰完备,适应性和泛化能力良好的优点, 核心思想是在特征空间中寻找到一个最优超平面将两类样本尽可能大的分开, 能够较好的处理小样本、非线性和克服“维数灾难”问题, 并且表现出优秀的分类能力和泛化能力而被广泛应用于分类和回归等领域. 但是 SVM 对核函数的参数选取对分类效果影响很大, 不合适的参数可能使得分类器性能大大降低. 针对 SVM 核参数的选取问题, 目前尚没有统一有效的方法. 传统的参数选择方法如实验法、网格搜索法等由于耗时过长和不必要的验证流程等缺点, 更常用的方法是群智能算法如蚁群算法、遗传算法和粒子群算法等优化支持向量机核参数.粒子群算法由于算法结构简单、寻优能力相对较好, 近年来选择哈里斯鹰算法优化 SVM 参数成为研究热点之一.。2 部分代码%% function [Rabbit_Energy,Rabbit_Location,CNVG] NCHHO_IoV(N,T,lb,ub,dim,fobj) % initialize the location and Energy of the rabbit Rabbit_Locationzeros(1,dim); Rabbit_Energy0; %Initialize the locations of Harris hawks Xinitialization(N,dim,ub,lb); CNVGzeros(1,T); t0; % Loop counter while tT for i1:size(X,1) % Check boundries FUX(i,:)ub;FLX(i,:)lb;X(i,:)(X(i,:).*(~(FUFL)))ub.*FUlb.*FL; % fitness of locations fitnessfobj(X(i,:)); % Update the location of Rabbit if fitnessRabbit_Energy Rabbit_Energyfitness; Rabbit_LocationX(i,:); end end E1abs(2*(1-(t/T))-2); % factor to show the decreaing energy of rabbit a1 4; % Initial chaotic map parameter configuration teta 0.7; % Initial chaotic map parameter configuration % Update the location of Harris hawks for i1:size(X,1) for ii1:4 Cm(1,ii) abs((a1/4)*sin(pi*teta)); teta Cm(1,ii); end E02*rand()-1; %-1E01 Escaping_EnergyE1*(E0); % escaping energy of rabbit if abs(Escaping_Energy)1 %% Exploration: % Harris hawks perch randomly based on 2 strategy: qrand(); rand_Hawk_index floor(N*rand()1); X_rand X(rand_Hawk_index, :); if q0.5 % perch based on other family members X(i,:)X_rand-Cm(1,1)*abs(X_rand-2*Cm(1,2)*X(i,:)); elseif q0.5 % perch on a random tall tree (random site inside groups home range) X(i,:)(Rabbit_Location(1,:)-mean(X))-Cm(1,3)*((ub-lb)*Cm(1,4)lb); end elseif abs(Escaping_Energy)1 %% Exploitation: % Attacking the rabbit using 4 strategies regarding the behavior of the rabbit %% phase 1: surprise pounce (seven kills) % surprise pounce (seven kills): multiple, short rapid dives by different hawks rrand(); % probablity of each event if r0.5 abs(Escaping_Energy)0.5 % Hard besiege X(i,:)(Rabbit_Location)-Escaping_Energy*abs(Rabbit_Location-X(i,:)); end if r0.5 abs(Escaping_Energy)0.5 % Soft besiege Jump_strength2*(1-rand()); % random jump strength of the rabbit X(i,:)(Rabbit_Location-X(i,:))-Escaping_Energy*abs(Jump_strength*Rabbit_Location-X(i,:)); end %% phase 2: performing team rapid dives (leapfrog movements) if r0.5 abs(Escaping_Energy)0.5 % Soft besiege % rabbit try to escape by many zigzag deceptive motions w12*exp(-(8*t/T)^2); % Non-linear control Parameter Jump_strength2*(1-rand()); X1w1*Rabbit_Location-Escaping_Energy*abs(Jump_strength*Rabbit_Location-X(i,:)); if fobj(X1)fobj(X(i,:)) % improved move? X(i,:)X1; else % hawks perform levy-based short rapid dives around the rabbit X2w1*Rabbit_Location-Escaping_Energy*abs(Jump_strength*Rabbit_Location-X(i,:))rand(1,dim).*Levy(dim); if (fobj(X2)fobj(X(i,:))) % improved move? X(i,:)X2; end end end if r0.5 abs(Escaping_Energy)0.5 % Hard besiege % rabbit try to escape by many zigzag deceptive motions % hawks try to decrease their average location with the rabbit w12*exp(-(8*t/T)^2); Jump_strength2*(1-rand()); X1w1*Rabbit_Location-Escaping_Energy*abs(Jump_strength*Rabbit_Location-mean(X)); if fobj(X1)fobj(X(i,:)) % improved move? X(i,:)X1; else % Perform levy-based short rapid dives around the rabbit X2w1*Rabbit_Location-Escaping_Energy*abs(Jump_strength*Rabbit_Location-mean(X))rand(1,dim).*Levy(dim); if (fobj(X2)fobj(X(i,:))) % improved move? X(i,:)X2; end end end %% end end tt1; CNVG(t)Rabbit_Energy; end end % ___________________________________ function oLevy(d) beta1.5; sigma(gamma(1beta)*sin(pi*beta/2)/(gamma((1beta)/2)*beta*2^((beta-1)/2)))^(1/beta); urandn(1,d)*sigma;vrandn(1,d);stepu./abs(v).^(1/beta); ostep; end3 仿真结果4 参考文献[1]颜晓娟. 基于改进遗传算法寻优的SVM风能短期预测. Diss. 广西大学, 2015.部分理论引用网络文献若有侵权联系博主删除。
