AI_v3

import pandas as pd import numpy as np dataset = pd.read_csv('./churn.csv') # Local File dataset = dataset[dataset.iloc[:,0:18]!= 'Unknown'] #drop all rows where status is unknown dataset = dataset.dropna(axis=0, how='any') #drop na rows # set x and y values X = dataset.iloc[:, 3:18].values y = dataset.iloc[:, 2].values

# Encoding categorical data from sklearn import * from sklearn.preprocessing import LabelEncoder, OneHotEncoder from sklearn.metrics import ConfusionMatrixDisplay labelencoder = LabelEncoder() for i in range(1,7): if i != 2: X[:, i] = labelencoder.fit_transform(X[:, i]) # Y encoding 1 = Existing Customer / 0 = Attrited Customer y[:] = labelencoder.fit_transform(y[:]) from sklearn.compose import ColumnTransformer ct = ColumnTransformer([("X", OneHotEncoder(dtype = int), [1])], remainder = 'passthrough') X = ct.fit_transform(X) X = X[:, 1:] print(X) # Splitting the dataset into the Training set and Test set from sklearn.model_selection import train_test_split X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.2, random_state = 0) # Feature Scaling from sklearn.preprocessing import StandardScaler sc = StandardScaler() X_train = sc.fit_transform(X_train) X_test = sc.transform(X_test)

# Importing the Keras libraries and packages from tensorflow.keras import * from tensorflow.keras.models import Sequential from tensorflow.keras.layers import Dense from tensorflow.keras import layers from tensorflow.keras.callbacks import Callback import tensorflow as tf # Initialising the ANN classifier = Sequential( [ layers.Dense(24, kernel_initializer = 'uniform', activation = 'relu'), layers.Dense(24, kernel_initializer = 'uniform', activation = 'relu'), layers.Dense(24, kernel_initializer = 'uniform', activation = 'relu'), layers.Dense(12, kernel_initializer = 'uniform', activation = 'relu'), layers.Dense(1, kernel_initializer = 'uniform', activation = 'sigmoid') ] ) # Compiling the ANN classifier.compile(optimizer = 'adam', loss = 'binary_crossentropy', metrics = ['accuracy']) # Testing X_t = np.asarray(X_train).astype('float32').reshape(-1,15) Y_t = np.asarray(y_train).astype('float32').reshape(-1,1) class Callback(tf.keras.callbacks.Callback): def __init__(self): self.data = [] def on_epoch_end(self,epoch,logs): self.data.append(logs) def data(self): return self.data batch_size = 32 callback = Callback() data = classifier.fit( X_t, Y_t, batch_size = batch_size, # Default epochs = int(len(X_t) / batch_size), callbacks=[callback], verbose=1 ) #print(data.params)

# Part 3 - Making the predictions and evaluating the model # Predicting the Test set results y_pred = classifier.predict(X_test) y_pred = (y_pred > 0.5) # Making the Confusion Matrix from sklearn.metrics import confusion_matrix y_pred2 = [int(i) for i in y_pred] y_test2 = list(y_test) # Confusion Matrix Plot ConfusionMatrixDisplay.from_predictions(y_test2, y_pred2) # Computing Accuracy, Precision and Recall from sklearn.metrics import accuracy_score, precision_score, recall_score print("Accuracy :",accuracy_score(y_test2,y_pred2)) print("Precision :",precision_score(y_test2,y_pred2)) print("Recall :",recall_score(y_test2,y_pred2))

import matplotlib.pyplot as plt x = [] y = [] for i in callback.data: x.append(i["loss"]) y.append(i["accuracy"]) fig, ax = plt.subplots(constrained_layout=True) pl1 = ax.plot(range(0,len(y)) , y,label='Accuracy') ax.set_xlabel('Epoch') ax2 = ax.twinx() pl2 = ax2.plot(range(0,len(x)) , x , color='#ff6f00',label='Loss Function') lns = pl1+pl2 labs = [l.get_label() for l in lns] ax.legend(lns, labs, loc='center right') fig.tight_layout() # otherwise the right y-label is slightly clipped plt.show()