mnist.py

# -*- coding: utf-8 -*-
"""MNIST.ipynb

Automatically generated by Colaboratory.

Original file is located at
    https://colab.research.google.com/drive/12a1YXrUCfIgib4Fke02qj5qE5y6tJg_h
"""

import keras
import numpy as np
from keras import datasets, layers, models

from keras.datasets import mnist
from matplotlib import pyplot

#loading
(train_X, train_y), (test_X, test_y) = mnist.load_data()
 
#shape of dataset
print('X_train: ' + str(train_X.shape))
print('Y_train: ' + str(train_y.shape))
print('X_test:  '  + str(test_X.shape))
print('Y_test:  '  + str(test_y.shape))
 
#plotting
from matplotlib import pyplot
for i in range(9):  
  pyplot.subplot(330 + 1 + i)
  pyplot.imshow(train_X[i], cmap=pyplot.get_cmap('gray'))
  pyplot.show()

train_X = train_X / 255
test_X = test_X / 255

train_X = train_X.reshape(-1,28,28,1)
test_X = test_X.reshape(-1,28,28,1)

convolutional_neural_network = models.Sequential([
    layers.Conv2D(filters=25, kernel_size=(3, 3), activation='relu', input_shape=(28,28,1)), 
    layers.MaxPooling2D((2, 2)),
    layers.Conv2D(filters=64, kernel_size=(3, 3), activation='relu'),
    layers.MaxPooling2D((2, 2)),
    layers.Conv2D(filters=64, kernel_size=(3, 3), activation='relu'),
    layers.MaxPooling2D((2, 2)),
    layers.Flatten(),
    layers.Dense(64, activation='relu'),
    layers.Dense(10, activation='softmax')
])

convolutional_neural_network.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['accuracy'])
convolutional_neural_network.fit(train_X, train_y, epochs=10)

convolutional_neural_network.evaluate(test_X, test_y)
y_predicted_by_model = convolutional_neural_network.predict(test_X)
y_predicted_by_model[0]
np.argmax(y_predicted_by_model[0])