Machine Learning Libraries and Functions Used in Assignment

1 Numpy and some python

scipy reading an image from file and return a numpy ndarray(N-dimensional array)

ndimage.imread(image_file)

Convert the data type in the numpy ndarray to desired type

ndimage.imread(image_file).astype(float)

Mathematic operation to every element in the vector or matrix

ndimage.imread(image_file).astype(float) - pixel_depth/2

Numpy index slicing, since there are 3 dims, the following is selecting the first dim from 0 to num_images-1

dataset = dataset[0:num_images, :, :]

Python pickle dump/serialize data to file

#Serializing to a file
pickle.dump(dataset, f, pickle.HIGHEST_PROTOCOL)
#De-serializing from a file
with open(pickle_file, 'rb') as f:
  letter_set = pickle.load(f)

Making numpy N-dimensional arrarys

dataset = np.ndarray((nb_rows, img_size, img_size), dtype=np.float32
labels = np.ndarray(nb_rows, dtype=np.int32)

Two methods random shuffle of the data

#Method 1 letter_set is numpy ndarray
np.random.shuffle(letter_set)
#Method 2 Using index slicing to randomize the data
permutation = np.random.permutation(labels.shape[0])
shuffled_dataset = dataset[permutation,:,:]