Laboratorio Matlab: differenze tra le versioni

Da Bioingegneria Elettronica e Informatica.

Vai a: navigazione, ricerca

Versione delle 15:21, 21 mag 2019

Reti Neurali Feed-Forward

Esempio di applicazione di reti neurali artificiali.

Creazione dataset:

```
load cancer_dataset.mat
```
```
x = cancerInputs;
```
```
t = cancerTargets(1,:);
```
```
 
```
```
temp = [x;t];
```
```
rng(0)
```
```
p = randperm(size(temp,2));
```
```
 
```
```
train_size = floor(size(temp,2)*.8);
```
```
p_train = p(1:train_size);
```
```
p_test = p(train_size+1:end);
```
```
 
```
```
x_train = temp(1:9,p_train);
```
```
t_train = temp(10,p_train);
```
```
 
```
```
x_test = temp(1:9,p_test);
```
```
t_test = temp(10,p_test);
```
```
 
```

save('dataset2.mat','x_train','t_train','x_test','t_test')

Implementazione rete neurale:

```
load dataset2.mat
```
```
 
```

disp(['# Features: ',num2str(size(x_train,1))])

disp(['# Samples: ',num2str(size(x_train,2))])

```
 
```
```
%% Creazione rete
```
```
 
```
```
% Layers nascosti
```
```
% hiddenLayerSize = [20];
```
```
% hiddenLayerSize = [50];
```
```
hiddenLayerSize = [20,10];
```
```
 
```
```
 
```
```
% Training Function - help nntrain
```

trainFcn = 'traingdx'; % traingda, traingdm, traingd

```
 
```
```
% Creazione rete
```

net = patternnet(hiddenLayerSize, trainFcn);

```
 
```
```
% Suddivisione dataset
```
```
net.divideFcn = 'dividerand';
```
```
net.divideParam.trainRatio = 70/100;
```
```
net.divideParam.valRatio = 30/100;
```
```
net.divideParam.testRatio = 0/100;
```
```
 
```
```
% Criteri di stop
```
```
net.trainParam.epochs = 5000;;
```
```
%net.trainParam.max_fail = 20;
```
```
%net.trainParam.min_grad = 0;%10e-5;
```
```
 
```
```
% Funzione errore
```
```
net.performFcn = 'mse';
```
```
 
```
```
% Funzioni di attivazione
```
```
net.layers{end}.transferFcn = 'logsig';
```
```
 
```
```
% Visualizza rete
```
```
view(net)
```
```
 
```
```
%% Inizializzazione Rete
```
```
rng(0)
```
```
net = configure(net,x_train,t_train);
```
```
net = init(net);
```
```
init_LW = net.LW;
```
```
init_IW = net.IW;
```
```
 
```
```
%% Addestramento Rete
```
```
[net,tr] = train(net,x_train,t_train);
```
```
y_train = net(x_train);
```
```
 
```
```
% Plots vari
```
```
figure, plotperform(tr)
```
```
%figure, plottrainstate(tr)
```
```
%figure, ploterrhist(e)
```

figure, plotconfusion(t_train,y_train),title('Training Confusion Matrix')

```
%figure, plotroc(t,y)
```
```
 
```
```
%% Test Rete
```
```
y_test = net(x_test);
```

figure, plotconfusion(t_test,y_test),title('Test Confusion Matrix')

Cluster Analysis: Self-Organizing Maps

Implementazione rete SOM:

```
load filteredyeastdata.mat
```
```
rng(0);
```

[x,std_settings] = mapstd(yeastvalues');  % Normalize data

[x,pca_settings] = processpca(x,0.15);    % PCA

```
 
```
```
clusters_number = 10;
```
```
 
```
```
dimensions = [clusters_number];
```
```
net = selforgmap(dimensions);
```
```
view(net)
```
```
 
```
```
net.trainParam.epochs = 2000;
```
```
net.trainParam.showCommandLine = 1;
```
```
 
```
```
net = train(net,x);
```
```
 
```
```
figure,plotsompos(net,x);
```
```
figure,plotsomhits(net,x);
```
```
figure,plotsomnd(net,x);
```
```
 
```
```
y = net(x);
```
```
cluster_indices = vec2ind(y);
```

Dendrogramma:

```
range = 1:100;
```

cgo = clustergram(yeastvalues(range,:),'RowLabels',genes(range),'ColumnLabels',times,'Standardize','Row');

```
get(cgo)
```

Estratto da "http://www.vitoantoniobevilacqua.it/wiki/index.php?title=Laboratorio_Matlab&oldid=945"