🔥알림🔥
① 테디노트 유튜브 -
구경하러 가기!
② LangChain 한국어 튜토리얼
바로가기 👀
③ 랭체인 노트 무료 전자책(wikidocs)
바로가기 🙌
Kaggle Digit Recognizer Competition - achieve over 99% accuracy!
This is a very simple tutorial for tensorflow 2.0 beginners!
This tutorial is based on Digit Recognizer presented by Kaggle
You may be able to submit with the output generated by your model
Goal
- using tensorflow 2.0 frameworks
- using Sequential Models
- applying Image Augmentation by using ImageDataGenerator
- Make submission to Kaggle
Import Libraries
In [1]:
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import warnings
warnings.filterwarnings('ignore')
%matplotlib inline
Load Data
In [2]:
train = pd.read_csv('data/train.csv')
test = pd.read_csv('data/test.csv')
In [3]:
x_train = train.drop('label', 1)
y_train = train['label']
Explore class distributions
In [4]:
y_train.value_counts()
Out[4]:
In [5]:
x_train.shape, y_train.shape
Out[5]:
Reshaping Model to 28 * 28
change shape to (28, 28, 1)
28 X 28 picture size with graysclae channel of 1
In [6]:
x_train = x_train.values.reshape(-1, 28, 28, 1)
In [7]:
y_train = y_train.values.reshape(-1, 1)
In [8]:
y_train.shape
Out[8]:
x_test refers to test datasets that I need to predict
In [9]:
x_test = test.values.reshape(-1, 28, 28, 1)
In [10]:
x_test.shape
Out[10]:
Using ImageDataGenerator to augment Images
In [11]:
from tensorflow.keras.preprocessing.image import ImageDataGenerator
In [12]:
datagen = ImageDataGenerator(
rescale=1.0/255,
featurewise_center=False, # set input mean to 0 over the dataset
samplewise_center=False, # set each sample mean to 0
featurewise_std_normalization=False, # divide inputs by std of the dataset
samplewise_std_normalization=False, # divide each input by its std
zca_whitening=False, # apply ZCA whitening
rotation_range=10, # randomly rotate images in the range (degrees, 0 to 180)
zoom_range = 0.1, # Randomly zoom image
width_shift_range=0.1, # randomly shift images horizontally (fraction of total width)
height_shift_range=0.1, # randomly shift images vertically (fraction of total height)
horizontal_flip=False, # randomly flip images
vertical_flip=False) # randomly flip images
In [13]:
from sklearn.model_selection import train_test_split
In [14]:
x_train, x_val, y_train, y_val = train_test_split(x_train, y_train, test_size=0.2)
In [15]:
x_train.shape, x_val.shape
Out[15]:
Modeling
In [16]:
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense, Conv2D, MaxPool2D, Flatten, Dropout
from tensorflow.keras.callbacks import ModelCheckpoint
In [17]:
model = Sequential()
model.add(Conv2D(filters=32, kernel_size=(5, 5), padding='same', activation='relu', input_shape=(28, 28, 1)))
model.add(Conv2D(filters=32, kernel_size=(5, 5), padding='same', activation='relu', input_shape=(28, 28, 1)))
model.add(MaxPool2D(pool_size=(2, 2), strides=2))
model.add(Conv2D(filters=64, kernel_size=(3, 3), padding='same', activation='relu'))
model.add(Conv2D(filters=64, kernel_size=(3, 3), padding='same', activation='relu'))
model.add(MaxPool2D(pool_size=(2, 2), strides=2))
model.add(Dropout(rate=0.25))
model.add(Flatten())
model.add(Dense(1024, activation='relu'))
model.add(Dropout(rate=0.25))
model.add(Dense(10, activation='softmax'))
In [18]:
model.summary()
In [19]:
batch_size = 128
epochs = 20
In [20]:
model_path = 'model/tmp-best-model.h5'
In [21]:
checkpoint = ModelCheckpoint(model_path,
verbose=1,
monitor='val_loss',
save_best_only=True,
save_weight_only=True,
mode='auto')
In [22]:
model.compile(optimizer='adam', loss='sparse_categorical_crossentropy', metrics=['accuracy'])
In [23]:
history = model.fit(datagen.flow(x_train, y_train, batch_size=batch_size),
epochs = epochs, validation_data = (x_val, y_val),
steps_per_epoch=x_train.shape[0] // batch_size,
callbacks=[checkpoint],
)
Visualize
In [24]:
plt.figure(figsize=(12, 6))
plt.plot(np.arange(20)+1, history.history['loss'], label='Loss')
plt.plot(np.arange(20)+1, history.history['val_loss'], label='Validation Loss')
plt.title('losses over training', fontsize=20)
plt.xlabel('epochs', fontsize=15)
plt.ylabel('loss', fontsize=15)
plt.legend()
plt.show()
In [25]:
plt.figure(figsize=(12, 6))
plt.plot(np.arange(20)+1, history.history['accuracy'], label='Accuracy')
plt.plot(np.arange(20)+1, history.history['val_accuracy'], label='Validation Accuracy')
plt.title('Accuracy over training', fontsize=20)
plt.xlabel('epochs', fontsize=15)
plt.ylabel('Accuracy', fontsize=15)
plt.legend()
plt.show()
Load best model
In [26]:
model.load_weights(model_path)
Make Prediction
In [27]:
result = model.predict(x_test)
In [28]:
result = np.argmax(result, axis=1)
In [29]:
result = pd.Series(result, name='label')
Make Submission
In [30]:
submission = pd.read_csv('data/sample_submission.csv')
In [31]:
submission['Label'] = result
In [32]:
submission['Label'].value_counts()
Out[32]:
In [33]:
submission.to_csv('submission/final_submission.csv', index=False)
댓글남기기