In this chapter, you will explore the basics of machine learning methods used in marketing. You will learn about different types of machine learning, data preparation steps, and will run several end to end models to understand their power.

Why use ML for marketing? Strategies and use cases

Identify supervised learning examples

Supervised vs. unsupervised learning

Preparation for modeling

Investigate the data

Separate numerical and categorical columns

Encode categorical and scale numerical variables

ML modeling steps

Split data to training and testing

Fit a decision tree

Predict churn with decision tree

Machine learning for marketing basics

In this chapter you will learn churn prediction fundamentals, then fit logistic regression and decision tree models to predict churn. Finally, you will explore the results and extract insights on what are the drivers of the churn.

Churn prediction fundamentals

Explore churn rate and split data

Separate features and target variable

Predict churn with logistic regression

Fit logistic regression model

Fit logistic regression with L1 regularization

Identify optimal L1 penalty coefficient

Predict churn with decision trees

Fit decision tree model

Identify optimal tree depth

Identify and interpret churn drivers

Explore logistic regression coefficients

Break down decision tree rules

Churn prediction and drivers

In this chapter, you will learn the basics of Customer Lifetime Value (CLV) and its different calculation methodologies. You will harness this knowledge to build customer level purchase features to predict next month's transactions using linear regression.

Customer Lifetime Value (CLV) basics

Build retention and churn tables

Explore retention and churn

Calculating and projecting CLV

Calculate basic CLV

Calculate granular CLV

Calculate traditional CLV

Data preparation for purchase prediction

Build features

Define target variable

Predicting customer transactions

Predict next month transactions

Measure model fit

Explore model coefficients

Customer Lifetime Value (CLV) prediction

This final chapter dives into customer segmentation based on product purchase history. You will explore two different models that provide insights into purchasing patterns of customers and group them into well separated and interpretable customer segments.

Customer and product segmentation basics

Explore customer product purchase dataset

Understand differences in variables

Data preparation for segmentation

Unskew the variables

Normalize the variables

Build customer and product segmentation

Determine the optimal number of clusters

Build segmentation with k-means clustering

Alternative segmentation with NMF

Visualize and interpret segmentation solutions

K-means segmentation averages

NMF segmentation averages

Congratulations!

Customer segmentation

Telecom Dataset

The rise of machine learning (almost sounds like "rise of the machines"?) and applications of statistical methods to marketing have changed the field forever. Machine learning is being used to optimize customer journeys which maximize their satisfaction and lifetime value. This course will give you the foundational tools which you can immediately apply to improve your company’s marketing strategy. You will learn how to use different techniques to predict customer churn and interpret its drivers, measure, and forecast customer lifetime value, and finally, build customer segments based on their product purchase patterns. You will use customer data from a telecom company to predict churn, construct a recency-frequency-monetary dataset from an online retailer for customer lifetime value prediction, and build customer segments from product purchase data from a grocery shop.

Supervised Learning with scikit-learn

From customer lifetime value, predicting churn to segmentation - learn and implement Machine Learning use cases for Marketing in Python.

Machine Learning for Marketing in Python

Analyse marketing en Python

Head of Analytics & Science, Amazon

Introduction to logistic regression

Modeling steps

Fitting the model

Model performance metrics

Measuring model accuracy

Measuring precision and recall

Regularization

L1 regularization and feature selection

Tuning L1 regularization

Choosing optimal C value

Let's run some logistic regression models!

https://s3.amazonaws.com/assets.datacamp.com/production/course_19815/subtitles/course_19815_f2a60040b0e7218490069d6270c73957.vtt

Churn prediction and drivers - Predict churn with logistic regression

Predict churn with logistic regression

Create Your Free Account