|
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
! c9 |) Y7 a2 O; R- ?
NSGA2算法特征选择MATLAB实现(多目标) {. }# Z ~$ G, Y* d1 z# q# p
4 y5 w& k8 k- L' u
" i1 L* g* l% L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 [) j9 H- ^7 C! R: n! s
7 {* Y/ a4 n) V0 q% u) n3 m需要优化的两个目标为特征数和精度。; Z) `, H, d6 |9 K9 j; t
4 {5 u0 u! Q* f1 i' Tnsga2是一个多目标优化算法。
& Q5 z' K9 G3 u" O4 m; d x7 [) a! m0 L
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)
# a0 }, Z1 [2 {. T
# \! S: a7 X$ x0 D$ M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:$ K" t4 v- c+ a* w# ], L
0 {1 g, L+ D5 c* h/ L& ^- m# |
- 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./ a! i' w9 m+ Y7 |
! N) E( w3 b! L5 P6 a n3 a/ j
% l) J1 N2 [" o- c% r C" yMATLAB代码主函数:& [) R( h! F2 z' U
* C: v7 j! }$ k: f; U$ L
- %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
- %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 x! ?; m8 m0 l/ E
8 D; J2 Y' { P2 u2 ^评价函数(利用林志仁SVM进行训练):# u [/ E& N3 {+ B" S% v
0 p( d* v9 D( _5 s: \# L% D; P s( B- 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
, E/ h$ ~! i' u$ h8 O
7 N. o" c$ g4 M' B( n选的的数据集请从UCI上下载。. @/ g8 _- p7 N
2 j* L& i$ r* l2 C$ ]' X6 H, ?
结果:
/ W$ t# Q! S+ l) w* P4 {5 W' e/ b L$ q+ J
①pareto面
8 ]) y: F3 W. U6 T/ p
! L! x1 u" B% X* x
! Z7 W% |/ ]* y i B& q4 u: B& e( K( m$ F
最后粒子的数据(选出的特征数和精确度)
: N; V7 t7 j: X, m2 ?
1 d+ }1 \* b4 x8 R2 v8 F3 n: D* x
4 @& L1 {9 x# y7 s& i
3 v* x( P2 o! Q& G$ g9 W0 H% }2 \: V( h/ N
|
|