MNIST Data digit recognizer model.¶
Import required libraries to work with data and build our model¶
In [1]:
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.image as img
import seaborn as sns
%matplotlib inline
np.random.seed(0)
from sklearn.model_selection import train_test_split
from sklearn.metrics import confusion_matrix
import itertools
from tensorflow.keras.utils import to_categorical
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense, Dropout, Flatten, Conv2D, MaxPool2D
from tensorflow.keras.optimizers import RMSprop
from tensorflow.keras.preprocessing.image import ImageDataGenerator
from tensorflow.keras.callbacks import ReduceLROnPlateau
sns.set(style='white', context='notebook', palette='deep')
Load the data¶
In [2]:
train = pd.read_csv('train.csv')
test = pd.read_csv('test.csv')
Have a look at the data¶
In [3]:
train.head()
Out[3]:
In [4]:
test.head()
Out[4]:
In [5]:
y_train = train['label']
X_train = train.drop(['label'], axis=1)
sns.countplot(y_train)
y_train.value_counts()
Out[5]:
Normalizind the data before feeding in CNN, and also reshape the data¶
In [6]:
# Normalize the data
X_train = X_train / 255.0
test = test / 255.0
In [7]:
# Reshape image in 3 dimensions(height = 28px, width =28px, canal=1)
X_train = X_train.values.reshape(-1, 28, 28, 1)
test = test.values.reshape(-1, 28, 28, 1)
In [8]:
# Encode labels to one hot vectors
y_train = to_categorical(y_train, num_classes=10)
In [9]:
X_train, X_val, y_train, y_val = train_test_split(X_train, y_train, test_size=0.1, random_state=0)
Create Model¶
In [10]:
model = Sequential()
model.add(Conv2D(filters = 32, kernel_size = (5,5),
activation='relu', input_shape= (28,28,1)))
model.add(Conv2D(filters = 32, kernel_size = (5,5),
activation='relu', input_shape= (28,28,1)))
model.add(MaxPool2D(pool_size=(2,2)))
model.add(Dropout(0.25))
model.add(Conv2D(filters = 64, kernel_size = (3, 3),
activation='relu'))
model.add(Conv2D(filters = 64, kernel_size = (3,3),
activation='relu'))
model.add(MaxPool2D(pool_size=(2,2), strides=(2,2)))
model.add(Dropout(0.25))
model.add(Flatten())
model.add(Dense(256, activation = 'relu'))
model.add(Dropout(0.5))
model.add(Dense(10, activation='softmax'))
In [11]:
optimizer = RMSprop(lr=0.001, rho=0.9)
In [12]:
model.compile(optimizer=optimizer, loss='categorical_crossentropy', metrics=['accuracy'])
In [13]:
learning_rate_reduction = ReduceLROnPlateau(monitor='val_accuracy',
patience=3,
verbose=1,
factor=0.5,
min_lr=0.00001)
In [14]:
epochs = 30
batch_size = 86
In [15]:
datagen = ImageDataGenerator(
rotation_range=10,
zoom_range=0.1,
width_shift_range=0.1,
height_shift_range=0.1)
In [16]:
datagen.fit(X_train)
In [17]:
history = model.fit(datagen.flow(X_train, y_train, batch_size=batch_size),
epochs=epochs, validation_data=(X_val, y_val),
verbose=2, steps_per_epoch=X_train.shape[0]//batch_size,
callbacks=[learning_rate_reduction])
Pretty good accuracy with first model itself, let's create the submission file¶
In [18]:
results = model.predict(test)
results = np.argmax(results, axis=1)
results = pd.Series(results, name='Label')
In [19]:
sample_submission = pd.read_csv('sample_submission.csv')
In [20]:
sample_submission['Label'] = results
In [21]:
sample_submission.to_csv("submission.csv", index=False)
In [ ]: