Machine Learning for OpenCV 4 - Second Edition

Manually implementing bootstrapping in OpenCV

Bootstrapping can be implemented with the following procedure:

1. Load the dataset. Since we already did this earlier, we don't have to do it again.
2. Instantiate the classifier:

      In [15]: knn = cv2.ml.KNearest_create()      ...      knn.setDefaultK(1)

3. From our dataset with N samples, randomly choose N samples with replacement to form a bootstrap. This can be done most easily with the choice function from NumPy's random module. We tell the function to draw len(X) samples in the [0, len(X)-1] range with replacement (replace=True). The function then returns a list of indices, from which we form our bootstrap:

      In [16]: idx_boot = np.random.choice(len(X), size=len(X),      ...                                  replace=True) ... X_boot = X[idx_boot, ...