Artificial Intelligence Marketing and Predicting Consumer Choice by Struhl, Steven -- Read -- Imperial Library of Trantor

Index

Cover Title Page Copyright Contents List of Figures List of Tables Preface 01 Who should read this book and why?

What we cover in this book What can you expect in this book? Data versus information What is important? The methods we will be discussing Implicit views of people and biases One way of comparing these methods Sense and sensibility with predictions Where we will not be going Summary of key points

02 Getting the project going

At the beginning Know who you are talking about or talking to What is the most you can expect from each method? How do you judge the result? What is significant? On to correlations How do I plan to evaluate the results? Know what sensible goals might look like Summary of key points

03 Conjoint, discrete choice and other trade-offs: let’s do an experiment

The reasons we need these methods The basic thinking behind the experimentally designed methods What the methods ask – and get What is a designed experiment? The great measurement power of experiments Getting more from experiments: HB to the rescue A brief talk about origins Applications in brief Summary of key points

04 Creating the best, newest thing: discrete choice modelling

Key features Thinking through and setting up the problem How many people you need Utility and share Market simulations Making more than one choice: allocating purchases Using the simulator program in the online resources Rounding out the picture Summary of key points

05 Conjoint analysis and its uses

Thinking in conjoint versus thinking in choices Conjoint analysis for single-product optimization Using the single product simulator in the online resources Conjoint remains an excellent method for messages Conjoint analysis for the best service delivery Using the message optimization simulator in the online resources Conjoint analysis and interactions Variants of conjoint analysis Summary of key points

06 Predictive models: via classifications that grow on trees

Classification trees: understanding an amazing analytical method Seeing how trees work, step by step Strong, yet weak A case study: let’s take a cruise CHAID and CART (and CRT, C&RT, QUEST, J48 and others) Summary: applications and cautions

07 Remarkable predictive models with Bayes Nets

What are Bayes Nets and how do they compare with other methods? Let’s make a deal Our first example: Bayes Nets linking survey questions and behaviour Bayes Nets confirm a theoretical model, mostly What is important to buyers of children’s apparel Summary and conclusions

08 Putting it together: what to use when

The tasks the methods do Thinking about thinking

Bibliography Index Backcover