|
|
EDA365欢迎您登录!
您需要 登录 才可以下载或查看,没有帐号?注册
x
( T' `- f( f" o: C1 nNSGA2算法特征选择MATLAB实现(多目标) ) N/ V, Q& q6 ^/ U
5 T% t3 ]0 x$ w% C y+ x1 o
Z: x3 ]+ }" D0 l9 q利用nsga2进行进行特征选择其主要思想是:将子集的选择看作是一个搜索寻优问题(wrapper方法),生成不同的组合,对组合进行评价,再与其他的组合进行比较。这样就将子集的选择看作是一个是一个优化问题。
" x2 x) g8 W. M% [2 W
% e( C. u) {- Z需要优化的两个目标为特征数和精度。! E: M! _% m1 b5 g3 d# U) @
- J! q/ t8 J* _3 i8 Y; z
nsga2是一个多目标优化算法。+ r, Y, c, Q8 S' |
; V" h% m2 D2 E+ `6 c具体的nsga2通用算法请看:NSGA2算法MATLAB实现(能够自定义优化函数): h9 S! |8 E6 Y2 V
7 L8 V9 t, p3 A6 S' T1 D U6 j; D9 q
具体的特征选择代码在上述代码的基础上改了两个①主函数②评价函数,增加了一个数据分成训练集和测试集的函数:
7 L% E: j- \: t! \
( G. ~7 c4 e1 u0 y% j- N& @- function divide_datasets()
- load Parkinson.mat;
- dataMat=Parkinson_f;
- len=size(dataMat,1);
- %归一化
- maxV = max(dataMat);
- minV = min(dataMat);
- range = maxV-minV;
- newdataMat = (dataMat-repmat(minV,[len,1]))./(repmat(range,[len,1]));
- Indices = crossvalind('Kfold', length(Parkinson_label), 10);
- site = find(Indices==1|Indices==2|Indices==3);
- train_F = newdataMat(site,:);
- train_L = Parkinson_label(site);
- site2 = find(Indices~=1&Indices~=2&Indices~=3);
- test_F = newdataMat(site2,:);
- test_L =Parkinson_label(site2);
- save train_F train_F;
- save train_L train_L;
- save test_F test_F;
- save test_L test_L;
- end
- %what doesn't kill you makes you stronger, stand a little taller,doesn't mean i'm over cause you're gonw.
D' d5 {" a/ J; @& m7 l
: |% N* |+ o S) E N2 p3 s% K2 S8 e! p: ?
MATLAB代码主函数:; v/ G2 m% p( k0 _
/ N; S! a7 l- s* b2 Q
- %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
- %此处可以更改
- %更多机器学习内容请访问omegaxyz.com
- clc;
- clear;
- pop = 500; %种群数量
- gen = 100; %迭代次数
- M = 2; %目标数量
- V = 22; %维度
- min_range = zeros(1, V); %下界
- max_range = ones(1,V); %上界
- %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
- %特征选择
- divide_datasets();
- global answer
- answer=cell(M,3);
- global choice %选出的特征个数
- choice=0.8;
- %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
- chromosome = initialize_variables(pop, M, V, min_range, max_range);
- chromosome = non_domination_sort_mod(chromosome, M, V);
- for i = 1 : gen
- pool = round(pop/2);
- tour = 2;
- parent_chromosome = tournament_selection(chromosome, pool, tour);
- mu = 20;
- mum = 20;
- offspring_chromosome = genetic_operator(parent_chromosome,M, V, mu, mum, min_range, max_range);
- [main_pop,~] = size(chromosome);
- [offspring_pop,~] = size(offspring_chromosome);
- clear temp
- intermediate_chromosome(1:main_pop,:) = chromosome;
- intermediate_chromosome(main_pop + 1 : main_pop + offspring_pop,1 : M+V) = offspring_chromosome;
- intermediate_chromosome = non_domination_sort_mod(intermediate_chromosome, M, V);
- chromosome = replace_chromosome(intermediate_chromosome, M, V, pop);
- if ~mod(i,100)
- clc;
- fprintf('%d generations completed\n',i);
- end
- end
- if M == 2
- plot(chromosome(:,V + 1),chromosome(:,V + 2),'*');
- xlabel('f_1'); ylabel('f_2');
- title('Pareto Optimal Front');
- elseif M == 3
- plot3(chromosome(:,V + 1),chromosome(:,V + 2),chromosome(:,V + 3),'*');
- xlabel('f_1'); ylabel('f_2'); zlabel('f_3');
- title('Pareto Optimal SuRFace');
- end9 t0 a. C& ?3 c
6 ?7 x# ?. ?, W5 J7 C# g* v
评价函数(利用林志仁SVM进行训练):# n/ M% {3 r7 g, t
; `7 H* ^- s# j$ r7 g2 p6 f0 S
- function f = evaluate_objective(x, M, V, i)
- f = [];
- global answer
- global choice
- load train_F.mat;
- load train_L.mat;
- load test_F.mat;
- load test_L.mat;
- temp_x = x(1:V);
- inmodel = temp_x>choice;%%%%%设定恰当的阈值选择特征
- f(1) = sum(inmodel(1,:));
- answer(i,1)={f(1)};
- model = libsvmtrain(train_L,train_F(:,inmodel), '-s 0 -t 2 -c 1.2 -g 2.8');
- [predict_label, ~, ~] = libsvmpredict(test_L,test_F(:,inmodel),model,'-q');
- error=0;
- for j=1:length(test_L)
- if(predict_label(j,1) ~= test_L(j,1))
- error = error+1;
- end
- end
- error = error/length(test_L);
- f(2) = error;
- answer(i,2)={error};
- answer(i,3)={inmodel};
- end
" j- r ?1 C( _ 0 R9 e3 O# B8 G, P
选的的数据集请从UCI上下载。
& y2 g* M" L0 w2 _' w, W
: w Y. w) O' W* z: Z结果:- W+ d# H1 V! D4 a3 R/ e
* Y0 B2 J( v' W6 v! Z) ~
①pareto面1 O6 E" l# |! i# a( O
l$ E$ ?9 u& |& W2 ^3 J
, {) H! T/ |% @; S* o1 B
% |% z' G2 U/ j) o5 O最后粒子的数据(选出的特征数和精确度)
& @! v) U! j9 t% p _
- x# w+ A) l* f% [2 }- U5 f
. a9 n' E1 t8 ~0 ]% `$ y8 o; t2 @+ [& n7 n: I
4 h+ e3 R; j$ K0 ~: R) M: m |
|