eignFaces

Compute a PCA (eigenfaces) on the face dataset

n_samples: 760
n_features: 1850
n_classes: 3

Extracting the top 150 eigenfaces from 570 faces

done in 0.157s

Projecting the input data on the eigenfaces orthonormal basis

done in 0.017s

Calculationg PCA components' Variance:

Variance explained by the first principal component:  0.18954512666
Variance explained by the second principal component: 0.155786516991

PCA #components and f1_score relation - classification Report

n_components	x	precision	recall	f1-score	support
10	Donald Rumsfeld	0.54	0.61	0.57	36
15	Donald Rumsfeld	0.69	0.56	0.62	36
50	Donald Rumsfeld	0.97	0.86	0.91	36
150	Donald Rumsfeld	0.86	0.86	0.86	36
250	Donald Rumsfeld	0.83	0.83	0.83	36

Fitting the classifier to the training set

done in 3.839s

Best estimator found by grid search:

SVC(C=1000.0, cache_size=200, class_weight='balanced', coef0=0.0,
decision_function_shape=None, degree=3, gamma=0.001, kernel='rbf',
max_iter=-1, probability=False, random_state=None, shrinking=True,
tol=0.001, verbose=False)

Predicting the people names on the testing set

done in 0.010s

Classification Report

               | precision |   recall | f1-score  | support

------------------ | ----------| ----------|------------|------ Donald Rumsfeld | 0.91 | 0.83 | 0.87 | 36 George W Bush | 0.91 | 0.98 | 0.94 | 123 Gerhard Schroeder | 0.96 | 0.77 | 0.86 | 31

  avg / total   |    0.92   |   0.92  |    0.91    |   190

Confusion_matrix

[[ 30 6 0] [ 2 120 1] [ 1 6 24]]

Results