|
|
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
8 N: r% ~6 d! J9 `3 VNSGA2算法特征选择MATLAB实现(多目标)
( v+ D- \* u+ Z+ P' Y; H7 \: b3 G4 E) B$ ]! j1 s
# W! B8 a0 I% h; m9 S( |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。0 M( q3 N/ o* P/ \% [" A) T0 P' D
( L2 z( |. r' t
需要优化的两个目标为特征数和精度。
6 F: ^$ D' ]; B9 V4 ~( m: Z' t0 l6 F2 i9 X
nsga2是一个多目标优化算法。
7 F% q" @! v1 |% R1 T4 o
) D& Y5 ?/ R( ~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)
4 r1 z$ @; @& i0 R
3 w8 E" R% ~& j% X- D o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:
9 r i( o, V6 w2 C8 q
. I+ y7 E: L5 k" A- `& N- K- 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.: y4 R; N) h! x! G2 W
: P3 z ?4 @8 c; {- T. ?
0 Z1 h* o; k% ]+ n: w3 @MATLAB代码主函数:
' P9 p1 j6 s+ s3 `+ ?5 J
6 `+ n7 [& r0 T- %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
- %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');
- end
$ i. H( r) C' Z* R 8 |7 @1 L9 H5 f/ }
评价函数(利用林志仁SVM进行训练):
# ~* ], p. `4 d4 U: W6 l. C( I5 o( Z/ j0 T
- 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$ k9 R6 B. K1 ?$ X
! @* D6 f8 a0 o$ F* E3 M) L! E- K8 s
选的的数据集请从UCI上下载。
) A, V' U" z5 w8 r X$ X* T
1 A* M& p: P, P' D/ s+ ?" x% }: S4 I% W结果:
) p/ J" Y& Z9 h) r' {0 x2 w! v
5 X* ^1 N2 {0 [$ y. H) Q①pareto面. T5 V) ^4 ~( s$ j4 N& f# l5 X
3 \( W( D* K2 z# N
1 ]) n' _" G0 L6 N) b
1 i. g' j( t* n: @6 c# b* t: }, K
最后粒子的数据(选出的特征数和精确度)
# ?1 w" W+ p! L, h' F! Q
& ^4 q( S9 E# o+ |6 `1 n2 A% v/ H
- U, L2 v1 S! l- i: B. P6 {! Z/ P/ b. T. W s5 H7 f
|
|