Scribd
Upload
30 views

Coding

The document discusses building and evaluating a convolutional neural network (CNN) model for classifying handwritten Arabic characters. It loads and preprocesses a dataset of 13440 training and 3360 test images, normalizing the images and one-hot encoding the labels. A CNN is built with Conv2D, MaxPool2D, Flatten, and Dense layers and trained for 100 epochs with dropout and data augmentation. The model is evaluated on validation and test sets, achieving an accuracy score of X% on test data.

Uploaded by

Ajue Ramli
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
30 views

Coding

The document discusses building and evaluating a convolutional neural network (CNN) model for classifying handwritten Arabic characters. It loads and preprocesses a dataset of 13440 training and 3360 test images, normalizing the images and one-hot encoding the labels. A CNN is built with Conv2D, MaxPool2D, Flatten, and Dense layers and trained for 100 epochs with dropout and data augmentation. The model is evaluated on validation and test sets, achieving an accuracy score of X% on test data.

Uploaded by

Ajue Ramli
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

Link to the dataset:https://www.kaggle.

com/datasets/mloey1/ahcd1

import numpy as np
import os
import pandas as pd
import seaborn as sns
from PIL import Image
import matplotlib.pyplot as plt

import keras
from keras.models import Sequential
from keras.layers import Conv2D,MaxPool2D,Flatten,Dense,Dropout
from keras.utils import to_categorical
from keras.preprocessing.image import ImageDataGenerator

from google.colab import drive


drive.mount('/content/drive')

Mounted at /content/drive

Data Preprocessing

train_images = pd.read_csv('/content/drive/MyDrive/handwritten_arabic/csvTrainImages 13440x1024.csv')


train_label = pd.read_csv('/content/drive/MyDrive/handwritten_arabic/csvTrainLabel 13440x1.csv')

test_images = pd.read_csv('/content/drive/MyDrive/handwritten_arabic/csvTestImages 3360x1024.csv')


test_label = pd.read_csv('/content/drive/MyDrive/handwritten_arabic/csvTestLabel 3360x1.csv')

train_images = train_images.iloc[:,:].values
train_label = train_label.iloc[:,:].values

test_images = test_images.iloc[:,:].values
test_label = test_label.iloc[:,:].values

train_images[0].shape

train_label[0]

train_images.shape

train_images = train_images.reshape(train_images.shape[0], 32, 32)


test_images = test_images.reshape(test_images.shape[0],32,32)

print("Train images shape: ",train_images.shape)


print("Train labels shape: ",train_label.shape)

print("Test images shape: ",test_images.shape)


print("Test labels shape: ",test_label.shape)

fig, ax = plt.subplots(4,2,figsize=(20,20))

for i in range(4):
for j in range(2):
ax[i,j].imshow(train_images[i])

train_images.max()

train_images.min()

train_label.max()

#Normalize

train_images = train_images/255.0
test_images = test_images/255.0

train_images.max()
len(train_images)

X_train = train_images[1000:]
X_valid = train_images[:1000]

y_train = train_label[1000:]
y_valid = train_label[:1000]

num_classes = len(np.unique(train_label)) + 1
num_classes

X_test = test_images[:]
y_test = test_label[:]

y_train = to_categorical(y_train,num_classes)
y_valid = to_categorical(y_valid,num_classes)
y_test = to_categorical(y_test,num_classes)

y_train

print("X train shape: ",X_train.shape)

print("Training samples: ",X_train.shape[0])


print("Validation samples: ",X_valid.shape[0])
print("Testing samples: ",X_test.shape[0])

X_train = X_train.reshape([-1, 32, 32, 1])


X_test = X_test.reshape([-1, 32, 32, 1])
X_valid = X_valid.reshape([-1, 32, 32, 1])

X_train.shape

model = Sequential()

X_train.shape[1:]

Deep Learning Model

model.add(Conv2D(filters=16, kernel_size=2, padding='same', activation='relu', input_shape=(32,32,1)))


model.add(MaxPool2D(pool_size=2))

model.add(Conv2D(filters=32, kernel_size=2, padding='same', activation='relu'))


model.add(MaxPool2D(pool_size=2))

model.add(Conv2D(filters=64, kernel_size=2, padding='same', activation='relu'))


model.add(MaxPool2D(pool_size=2))

model.summary()

model.add(Flatten())

model.add(Dense(512,activation='relu'))
model.add(Dropout(0.3))

model.add(Dense(256,activation='relu'))
model.add(Dropout(0.3))

model.add(Dense(29,activation='softmax'))

model.summary()

model.compile(loss='categorical_crossentropy', optimizer='rmsprop',
metrics=['accuracy'])

data_gen = ImageDataGenerator(width_shift_range=0.1,
height_shift_range=0.1,
horizontal_flip=True)
data_gen.fit(X_train)

batch_size=128

history = model.fit_generator(data_gen.flow(X_train, y_train, batch_size=batch_size),


steps_per_epoch=X_train.shape[0] // batch_size,
epochs=100, verbose=2,
validation_data=(X_valid, y_valid),
validation_steps=X_valid.shape[0] // batch_size)

plt.style.use('dark_background')
fig, ax = plt.subplots(2, 2, figsize=(10, 10))
sns.lineplot(x=np.arange(1, 101), y=history.history.get('loss'), ax=ax[0, 0])
sns.lineplot(x=np.arange(1, 101), y=history.history.get('accuracy'), ax=ax[0, 1])
sns.lineplot(x=np.arange(1, 101), y=history.history.get('val_loss'), ax=ax[1, 0])
sns.lineplot(x=np.arange(1, 101), y=history.history.get('val_accuracy'), ax=ax[1, 1])
ax[0, 0].set_title('Training Loss vs Epochs')
ax[0, 1].set_title('Training Accuracy vs Epochs')
ax[1, 0].set_title('Validation Loss vs Epochs')
ax[1, 1].set_title('Validation Accuracy vs Epochs')
fig.suptitle('Basic CNN model', size=16)
plt.show()

score = model.evaluate(X_test, y_test, verbose=0)


print('\n', 'Test accuracy:', score[1]*100)
print('\n', 'Total test Loss:', score[0]*100)

error 0s completed at 10:18 PM

You might also like