classification_neural_net.md

Note: This is a generated markdown export from the Jupyter notebook file classification_neural_net.ipynb. You can also view the notebook with the nbviewer from Jupyter.

Classification with a neural network

%matplotlib inline
import tensorflow as tf

import seaborn as sns
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt

from sklearn import datasets, metrics, model_selection

digits = datasets.load_digits()

fig, axes = plt.subplots(nrows=1, ncols=10, figsize=(10, 3))
for ax, image, label in zip(axes, digits.images, digits.target):
    ax.set_axis_off()
    ax.imshow(image, cmap=plt.cm.gray_r)
    ax.set_title('%i' % label)

plt.figure()
plt.imshow(digits.images[0], cmap=plt.cm.gray_r)
plt.colorbar()
plt.grid(False)
plt.show()

target = digits.target
data = digits.images

print("min value: {}".format(np.amin(data)))
print("max value: {}".format(np.amax(data)))
print("shape: {}".format(np.shape(data)))

min value: 0.0
max value: 16.0
shape: (1797, 8, 8)

X_train, X_test, y_train, y_test = model_selection.train_test_split(
    data, target, test_size=0.5)


X_train = X_train.astype('float32') / 16.
X_test = X_test.astype('float32') / 16.

df_train = pd.DataFrame(y_train, columns=['target'])
df_train['type'] = 'train'

df_test = pd.DataFrame(y_test, columns=['target'])
df_test['type'] = 'test'

df_set = df_train.append(df_test)

_ = sns.countplot(x='target', hue='type', data=df_set)     

print('train samples:', len(X_train))
print('test samples', len(X_test))

train samples: 898
test samples 899

model = tf.keras.Sequential([
    tf.keras.layers.Flatten(input_shape=(8, 8)),
    tf.keras.layers.Dense(128, activation='relu'),
    tf.keras.layers.Dense(10)
])




model.compile(optimizer='adam',
              loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),
              metrics=['accuracy'])


model.summary()

Model: "sequential"
_________________________________________________________________
Layer (type)                 Output Shape              Param #   
=================================================================
flatten (Flatten)            (None, 64)                0         
_________________________________________________________________
dense (Dense)                (None, 128)               8320      
_________________________________________________________________
dense_1 (Dense)              (None, 10)                1290      
=================================================================
Total params: 9,610
Trainable params: 9,610
Non-trainable params: 0
_________________________________________________________________

%%time
history = model.fit(X_train, y_train, epochs=100, validation_split = 0.2, verbose=0)

CPU times: user 6.55 s, sys: 822 ms, total: 7.37 s
Wall time: 5.77 s

hist = pd.DataFrame(history.history)
hist['epoch'] = history.epoch
hist.tail()

	loss	accuracy	val_loss	val_accuracy	epoch
95	0.007926	1.0	0.097244	0.977778	95
96	0.007718	1.0	0.097566	0.972222	96
97	0.007589	1.0	0.098171	0.977778	97
98	0.007448	1.0	0.098217	0.977778	98
99	0.007221	1.0	0.098246	0.977778	99

def plot_loss(history):
  plt.plot(history.history['loss'], label='loss')
  plt.plot(history.history['val_loss'], label='val_loss')
  
  plt.xlabel('Epoch')
  plt.ylabel('Error')
  plt.legend()
  plt.grid(True)

plot_loss(history)

test_loss, test_acc = model.evaluate(X_test,  y_test, verbose=0)
print('Test loss:', test_loss)
print('Test accuracy:', test_acc)

Test loss: 0.11751686781644821
Test accuracy: 0.9688543081283569

probability_model = tf.keras.Sequential([model, 
                                         tf.keras.layers.Softmax()])

predicted = [np.argmax(x) for x in probability_model.predict(X_test)]

confusion_matrix = pd.DataFrame(metrics.confusion_matrix(y_test, predicted))
confusion_matrix

	0	1	2	3	4	5	6	7	8	9
0	80	0	0	0	2	0	1	0	0	0
1	0	83	0	0	0	0	2	0	1	0
2	0	1	78	0	0	0	0	0	0	0
3	0	0	0	92	0	1	0	0	0	0
4	0	0	0	0	105	0	0	0	1	0
5	0	1	0	0	0	90	1	0	0	2
6	0	1	0	0	0	1	86	0	0	0
7	0	0	0	0	1	0	0	83	0	0
8	0	8	0	1	0	0	0	0	79	0
9	0	1	0	0	0	1	0	1	0	95

_ = sns.heatmap(confusion_matrix, annot=True, cmap="Blues")

print("accuracy: {:.3f}".format(metrics.accuracy_score(y_test, predicted)))
print("precision: {:.3f}".format(metrics.precision_score(y_test, predicted, average='weighted')))
print("recall: {:.3f}".format(metrics.recall_score(y_test, predicted, average='weighted')))
print("f1 score: {:.3f}".format(metrics.f1_score(y_test, predicted, average='weighted')))

accuracy: 0.969
precision: 0.970
recall: 0.969
f1 score: 0.969