[SVM] machine learning support vector machine - Face Recognition

from sklearn.datasets import fetch_lfw_people
faces = fetch_lfw_people(min_faces_per_person=60)
print(faces.target_names)
print(faces.images.shape)

['Ariel Sharon' 'Colin Powell' 'Donald Rumsfeld' 'George W Bush'
 'Gerhard Schroeder' 'Hugo Chavez' 'Junichiro Koizumi' 'Tony Blair']
(1348, 62, 47)

fig, ax = plt.subplots(3, 5)
for i, axi in enumerate(ax.flat):
    axi.imshow(faces.images[i], cmap='bone')
    axi.set(xticks=[], yticks=[],
            xlabel=faces.target_names[faces.target[i]])

• The size of each figure is [62 × 47]
• Here we put each pixel as a feature, but this feature too much, with PCA dimension reduction

from sklearn.svm import SVC
#from sklearn.decomposition import RandomizedPCA
from sklearn.decomposition import PCA
from sklearn.pipeline import make_pipeline

pca = PCA(n_components=150, whiten=True, random_state=42)
svc = SVC(kernel='rbf', class_weight='balanced')
model = make_pipeline(pca, svc)

from sklearn.model_selection import train_test_split
Xtrain, Xtest, ytrain, ytest = train_test_split(faces.data, faces.target,
                                                random_state=40)

Grid search to select the parameter

from sklearn.model_selection import GridSearchCV
param_grid = {'svc__C': [1, 5, 10],
              'svc__gamma': [0.0001, 0.0005, 0.001]}
grid = GridSearchCV(model, param_grid)

%time grid.fit(Xtrain, ytrain)
print(grid.best_params_)

Wall time: 51.5 s
{‘svc__C’: 5, ‘svc__gamma’: 0.001}

model = grid.best_estimator_
yfit = model.predict(Xtest)
yfit.shape

(337,)

Check the forecast results

fig, ax = plt.subplots(4, 6)
for i, axi in enumerate(ax.flat):
    axi.imshow(Xtest[i].reshape(62, 47), cmap='bone')
    axi.set(xticks=[], yticks=[])
    axi.set_ylabel(faces.target_names[yfit[i]].split()[-1],
                   color='black' if yfit[i] == ytest[i] else 'red')
fig.suptitle('Predicted Names; Incorrect Labels in Red', size=14);

from sklearn.metrics import classification_report
print(classification_report(ytest, yfit,
                            target_names=faces.target_names))

• Accuracy (precision) = number of correct prediction (TP) / the number of correctly predicted (TP + FP)
• Recall (recall) = the number of correct prediction (TP) / the number of prediction (TP + FN)
• F1 = 2 the accuracy of the recall rate / (+ recall accuracy)

                   precision    recall  f1-score   support

     Ariel Sharon       0.50      0.50      0.50        16
     Colin Powell       0.69      0.81      0.75        54
  Donald Rumsfeld       0.83      0.85      0.84        34
    George W Bush       0.94      0.88      0.91       136
Gerhard Schroeder       0.72      0.85      0.78        27
      Hugo Chavez       0.81      0.72      0.76        18
Junichiro Koizumi       0.87      0.87      0.87        15
       Tony Blair       0.85      0.76      0.80        37

      avg / total       0.83      0.82      0.82       337

Chaos matrix (see who's more likely to be confused)

from sklearn.metrics import confusion_matrix
mat = confusion_matrix(ytest, yfit)
sns.heatmap(mat.T, square=True, annot=True, fmt='d', cbar=False,
            xticklabels=faces.target_names,
            yticklabels=faces.target_names)
plt.xlabel('true label')
plt.ylabel('predicted label');