Getting Started with RapidMiner Studio

So here you are, a business analyst or developer or jack-of-all-trades. Things are going well, but it's time to fine-tune your business. You've probably collected a lot of data about your customers — maybe you know five characteristics and maybe you know 50. Some customers have remained loyal and some have left (or churned) and some you just don't know. It costs a lot of money and manpower to attract new customers, so you'd really like to keep your current users to help pay for that. But how do you know who might leave? Who do you spend marketing dollars on to entice them to stay?

Enter RapidMiner Studio. Using RapidMiner's code-free platform, you can quickly and easily create analytics workflows, called processes, to determine who to target. In these processes you transform your data by connecting the building blocks or operators that will eventually result in a model that helps you predict the future. Sound like fantasy or science fiction? It's not, it's modern analytics.

Getting down to business

This Getting Started tutorial will help familiarize you with some basic features and concepts of RapidMiner Studio. Using the enclosed data set and RapidMiner Studio, you will build a model that identifies the relationship between attributes (the characteristics of a customer) and labels (also known as target attributes).

Your use case: You are a mobile app company trying to maximize customer retention (minimize churn). App use requires registration; when a customer deletes the registration or the app, you classify them as having churned.

Each row in the enclosed data represents a user and provides the gender, age, payment method, and last transaction. (The last transaction is the number of weeks before now that the customer made a purchase — an app, a subscription, in-app purchases, etc.) In many rows the label is known (whether a customer churned or is loyal); in some it is not. Using this tutorial, you will build a model that predicts the label for each customer in which loyalty status is unknown.

The tutorial is broken into five parts:

Each part in the tutorial describes an important component in the design process, and each builds on the part(s) that came before. When you have finished all parts, you will have constructed a fully functional model that you can use as a starting point for your own real-world business models.

Additional Resources

Some additional resources available to help get up and running with RapidMiner Studio:

Resource Description
Getting Started glossary Quick descriptions of common RapidMiner and data mining terms used in this tutorial
User interface overview RapidMiner Studio interface cheat sheet
Repository samples Additional exercises for familiarizing yourself with RapidMiner Studio (once you have completed this tutorial)

Welcome to RapidMiner Studio, and on to Part 1!