Gain insights into applied machine learning for marketing analytics. Explore data science techniques, create predictive models with Python libraries, and drive results with data-driven decisions.

Marketing Analytics with MAchine Learning-b.png

analyticsoul_docker.tar.gz

00_helloworld

Notebook Live 3

In this course, you’ll learn applied machine learning in marketing analytics and cover modern data science techniques such as data exploration, data preprocessing, feature engineering and evaluation. You’ll gain hands-on experience with the Python libraries pandas, Scikit-learn, and seaborn, and learn how to use them to perform data wrangling, data analysis, create predictive models, and visualize your results.

This course will introduce you to basic data manipulation techniques. Further, you’ll cover specific topics such as customer revenue prediction using Linear Regression, customer segmentation using the K-Means Algorithm, customer churn prediction using Logistic Regression, and Customer Lifetime Value (CLV) analysis and prediction.

Whether you're a marketer or business professional, this course will give you the skills to make data-driven decisions and drive results. You’ll have hands-on experience predicting future revenue, segmenting your customers into different groups, and predicting customer churn

Marketing Analytics Using Machine Learning Techniques

# Task 1.1: Load the cdnow dataset and convert the date datatype
# Write your code here
df_cdnow = pd.read_csv('cdnow.csv', index_col=0)
df_cdnow['date'] = pd.to_datetime(df_cdnow['date'])

# Perform task 1.2 and 1.3 together with groupby() function
# Task 1.2: Identify first and last transaction date, also number of transactions
# Task 1.3: Aggregate total transaction value
# Write your code here
df_cdnow = df_cdnow.groupby('customer_id').agg({
    'date': ['min', 'max', 'nunique'],
    'price': 'sum'
})

# Task 1.4: Rename columns for better manage
# Write your code here
df_cdnow.columns = ['_'.join(col).strip() for col in df_cdnow.columns.values]
df_cdnow.rename(columns={
    'date_min': 'first_transaction',
    'date_max': 'last_transaction',
    'date_nunique': 'frequency',
    'price_sum': 'revenue'
}, inplace=True)

print(df_cdnow.head())


import pandas as pd
import numpy as np
import datetime as dt
import matplotlib.pyplot as plt
import seaborn as sns

from sklearn.preprocessing import StandardScaler
from scipy.cluster.hierarchy import dendrogram, linkage
from sklearn.cluster import KMeans


# Task 1: Read the dataset 'ad_clicks.csv'
df_adclicks = pd.read_csv('ad_clicks.csv')

# Task 2: Select the following features for your model
#         Age, Gender, Income, Internet Usage, Time Spent
X = df_adclicks[['Age', 'Gender', 'Income', 'Internet Usage', 'Time Spent']]

# Task 3: Select 'Is Clicked' as the target label
y = df_adclicks['Is Clicked']

# Task 4: Split training and testing datasets (features and labels)
#         Use 30% of the dataset for testing
#         Use random_state=123 to have the same result as the solution
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=123)

# Task 5: Instantiate the model with max_iter=1000
#         And train your model with the training dataset
logreg = LogisticRegression(max_iter=1000)
logreg.fit(X_train, y_train)

# Task 6: Predict model performance with testing dataset
y_pred = logreg.predict(X_test)

# Task 7: Calculate the accuracy, precision, and recall of your model
accuracy = accuracy_score(y_test, y_pred)
precision = precision_score(y_test, y_pred)
recall = recall_score(y_test, y_pred)

print('Accuracy:',  round(accuracy, 4))
print('Precision:', round(precision, 4))
print('Recall:',    round(recall, 4))

Walk-through the solution to the Customer Lifetime Value challenge.

Introduction

Data Manipulation

Predicting Customer Revenue

Customer Segmentation

Predicting Customer Churn

Predicting Customer Lifetime Value (CLV)

Conclusion

Solution: Customer Lifetime Value

Task 1