Scribd
Upload
21 views6 pages

DL 3 Ks

The document describes building an image classification model on the MNIST dataset in 4 stages: 1) Loading and preprocessing the image data, 2) Defining the model architecture as a convolutional neural network, 3) Training the model for 10 epochs, 4) Estimating the model performance on the test data, achieving over 98% accuracy.

Uploaded by

Priyanshu Agrawal
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
21 views6 pages

DL 3 Ks

The document describes building an image classification model on the MNIST dataset in 4 stages: 1) Loading and preprocessing the image data, 2) Defining the model architecture as a convolutional neural network, 3) Training the model for 10 epochs, 4) Estimating the model performance on the test data, achieving over 98% accuracy.

Uploaded by

Priyanshu Agrawal
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 6

[1]: #Name: kasturi Desai

#Roll No: 15
#Batch: B1 [IT]

[3]: #Build the Image classification model by dividing the model into following 4␣
↪stages:

#1.Loading and preprocessing the image data


#2.Defining the model's architecture
#3.Training the model
#4.Estimating the model's performance

[6]: import numpy as np


import pandas as pd
import random
import tensorflow as tf
import matplotlib.pyplot as plt
from sklearn.metrics import accuracy_score
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Flatten, Conv2D, Dense, MaxPooling2D
from tensorflow.keras.optimizers import SGD
from tensorflow.keras.utils import to_categorical
from tensorflow.keras.datasets import mnist

[8]: # Loading and preprocessing the image data


(X_train, y_train), (X_test, y_test) = mnist.load_data()

[9]: print(X_train.shape)
(60000, 28, 28)

(60000, 28, 28)

[9]: (60000, 28, 28)

[10]: X_train[0].min(), X_train[0].max()

[10]: (0, 255)

1
[11]: X_train = (X_train - 0.0) / (255.0 - 0.0)
X_test = (X_test - 0.0) / (255.0 - 0.0)
X_train[0].min(), X_train[0].max()
(0.0, 1.0)

[11]: (0.0, 1.0)

[18]: def plot_digit(image, digit, plt, i):


plt.subplot(4, 5, i + 1)
plt.imshow(image, cmap=plt.get_cmap('gray'))
plt.title(f"Digit: {digit}")
plt.xticks([])
plt.yticks([])
plt.figure(figsize=(16, 10))
for i in range(20):
plot_digit(X_train[i], y_train[i], plt, i)
plt.show()

[19]: X_train = X_train.reshape((X_train.shape + (1,)))


X_test = X_test.reshape((X_test.shape + (1,)))

[20]: y_train[0:20]

2
c=np.array([5, 0, 4, 1, 9, 2, 1, 3, 1, 4, 3, 5, 3, 6, 1, 7, 2, 8, 6,␣
↪9],dtype='u1')

[21]: # Defining the model's architecture


model = Sequential([
Conv2D(32, (3, 3), activation="relu", input_shape=(28, 28, 1)),
MaxPooling2D((2, 2)),
Flatten(),
Dense(100, activation="relu"),
Dense(10, activation="softmax")
])

[22]: optimizer = SGD(learning_rate=0.01, momentum=0.9)


model.compile(
optimizer=optimizer,
loss="sparse_categorical_crossentropy",
metrics=["accuracy"]
)

[23]:
model.summary()

Model: "sequential"

Layer (type) Output Shape Param #


=================================================================
conv2d (Conv2D) (None, 26, 26, 32) 320

max_pooling2d (MaxPooling2 (None, 13, 13, 32) 0


D)

flatten (Flatten) (None, 5408) 0

dense (Dense) (None, 100) 540900

dense_1 (Dense) (None, 10) 1010

=================================================================
Total params: 542230 (2.07 MB)
Trainable params: 542230 (2.07 MB)
Non-trainable params: 0 (0.00 Byte)

[37]: #Training and testing the model


Model_log=model.fit(X_train, y_train, epochs=10, batch_size=15,␣
↪verbose=1,validation_data=0);

Epoch 1/10

3
4000/4000 [==============================] - 26s 6ms/step - loss: 0.0044 -
accuracy: 0.9989
Epoch 2/10
4000/4000 [==============================] - 24s 6ms/step - loss: 0.0020 -
accuracy: 0.9996
Epoch 3/10
4000/4000 [==============================] - 27s 7ms/step - loss: 0.0018 -
accuracy: 0.9996
Epoch 4/10
4000/4000 [==============================] - 27s 7ms/step - loss: 8.3403e-04 -
accuracy: 0.9998
Epoch 5/10
4000/4000 [==============================] - 25s 6ms/step - loss: 4.2142e-04 -
accuracy: 0.9999
Epoch 6/10
4000/4000 [==============================] - 26s 7ms/step - loss: 1.9528e-04 -
accuracy: 1.0000
Epoch 7/10
4000/4000 [==============================] - 30s 7ms/step - loss: 1.3558e-04 -
accuracy: 1.0000
Epoch 8/10
4000/4000 [==============================] - 33s 8ms/step - loss: 1.0930e-04 -
accuracy: 1.0000
Epoch 9/10
4000/4000 [==============================] - 34s 9ms/step - loss: 9.6641e-05 -
accuracy: 1.0000
Epoch 10/10
4000/4000 [==============================] - 25s 6ms/step - loss: 8.5826e-05 -
accuracy: 1.0000
[41]: plt.figure(figsize=(16, 10))
for i in range(20):
image = random.choice(X_test).squeeze()
digit = np.argmax(model.predict(image.reshape((1, 28, 28, 1)))[0], axis=-1)
plot_digit(image, digit, plt, i)
plt.show()

1/1 [==============================] - 0s 16ms/step


1/1 [==============================] - 0s 17ms/step
1/1 [==============================] - 0s 17ms/step
1/1 [==============================] - 0s 18ms/step
1/1 [==============================] - 0s 17ms/step
1/1 [==============================] - 0s 17ms/step
1/1 [==============================] - 0s 18ms/step
1/1 [==============================] - 0s 18ms/step
1/1 [==============================] - 0s 16ms/step
1/1 [==============================] - 0s 17ms/step
1/1 [==============================] - 0s 17ms/step

4
1/1 [==============================] - 0s 17ms/step
1/1 [==============================] - 0s 20ms/step
1/1 [==============================] - 0s 23ms/step
1/1 [==============================] - 0s 20ms/step
1/1 [==============================] - 0s 19ms/step
1/1 [==============================] - 0s 21ms/step
1/1 [==============================] - 0s 21ms/step
1/1 [==============================] - 0s 18ms/step
1/1 [==============================] - 0s 20ms/step

[42]: predictions = np.argmax(model.predict(X_test), axis=-1)


accuracy_score(y_test, predictions)

313/313 [==============================] - 1s 3ms/step

[42]: 0.9896

[43]: n=random.randint(0,9999)
plt.imshow(X_test[n])
plt.show()

5
[44]: predicted_value=model.predict(X_test)
print("Handwritten number in the image is= %d" %np.argmax(predicted_value[n]))

313/313 [==============================] - 1s 3ms/step


Handwritten number in the image is= 0
[45]: # Estimating the model's performance
score = model.evaluate(X_test, y_test, verbose=0)
print('Test loss:', score[0])
print('Test accuracy:', score[1])

Test loss: 0.045085225254297256


Test accuracy: 0.9896000027656555
[ ]:

You might also like