Scribd
Upload
4 views5 pages

Deep Learning

The document outlines a Python script for building a neural network using Keras to predict customer churn based on data from 'Churn_Modelling.csv'. It includes steps for data preprocessing, encoding categorical variables, splitting the dataset, feature scaling, and constructing the artificial neural network (ANN) with specified layers and activation functions. Finally, it fits the model to the training data and evaluates its performance using a confusion matrix.

Uploaded by

yunusbahcesehir
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
4 views5 pages

Deep Learning

The document outlines a Python script for building a neural network using Keras to predict customer churn based on data from 'Churn_Modelling.csv'. It includes steps for data preprocessing, encoding categorical variables, splitting the dataset, feature scaling, and constructing the artificial neural network (ANN) with specified layers and activation functions. Finally, it fits the model to the training data and evaluates its performance using a confusion matrix.

Uploaded by

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

Deep Learning

Keras
PyTorch
TensorFlow
Theano
CNTK
Deeplearning4j
Caffe

Churn_Modelling.csv

Python neural_network.py

1 import pandas as pd
2
3 # Load the data
4 data = pd.read_csv('Churn_Modelling.csv')
5 X = [Link][:, 3:13].values
6 y = [Link][:, 13].values
7
8 # Encoding categorical data
9 from sklearn import preprocessing
10 le = [Link]()
11 X[:, 1] = le.fit_transform(X[:, 1])
12 le2 = [Link]()
13 X[:, 2] = le2.fit_transform(X[:, 2])
14
15 # One hot encoding
16 from [Link] import OneHotEncoder
17 from [Link] import ColumnTransformer
18 ct = ColumnTransformer([("ohe", OneHotEncoder(dtype=float),
[1])], remainder='passthrough')
19 X = ct.fit_transform(X)
20 X = X[:, 1:]
21
22 # Splitting the dataset into the Training set and Test set
23 from sklearn.model_selection import train_test_split
24 X_train, X_test, y_train, y_test = train_test_split(X, y,
test_size=0.33, random_state=0)
25
26 # Feature Scaling
27 from [Link] import StandardScaler
28 sc = StandardScaler()
29 X_train = sc.fit_transform(X_train)
30 X_test = [Link](X_test)
31
32 # Importing the Keras libraries and packages
33 from [Link] import Sequential # initialize the neural
network
34 from [Link] import Dense # build the layers
35
36 # Initialising the ANN
37 classifier = Sequential()
38
39 # Adding the input layer and the first hidden layer
40 [Link](Dense(units=6, kernel_initializer='uniform',
activation='relu', input_dim=11)) # units = 6 is the number
of nodes in the first hidden layer, input_dim = 11 is the
number inputs
41
42 # Adding the second hidden layer
43 [Link](Dense(units=6, kernel_initializer='uniform',
activation='relu')) # units = 6 is the number of nodes in
the second hidden layer
44
45 # Adding the output layer
46 [Link](Dense(units=1, kernel_initializer='uniform',
activation='sigmoid')) # units = 1 is the number of nodes
in the output layer, activation = 'sigmoid' is the
activation function for the output layer (sigmoid for
binary outcome).
47
48 # Compiling the ANN
49 [Link](optimizer='adam',
loss='binary_crossentropy', metrics=['accuracy']) #
optimizer = 'adam' is the algorithm to find the optimal
weights, loss = 'binary_crossentropy' is the loss function
within the stochastic gradient descent algorithm, metrics =
['accuracy'] is the metric used to evaluate the model.
50
51 # Fitting the ANN to the Training set
52 [Link](X_train, y_train, batch_size=10, epochs=100)
# batch_size = 10 is the number of observations after which
the weights are updated, epochs = 100 is the number of
times the whole dataset is passed through the ANN.
53
54 # Predicting the Test set results
55 y_pred = [Link](X_test)
56 y_pred = (y_pred > 0.5)
57
58 # Making the Confusion Matrix
59 from [Link] import confusion_matrix
60 cm = confusion_matrix(y_test, y_pred)
61 print(cm)
62 print((cm[0][0] + cm[1][1]) / len(y_test))

You might also like