Predictive Model Tutorial - Entering Your Data

Entering Your Predictive Model Data

Because the Predictive model requires a specific data format, users with their own data should review the pre-formatted template to learn about the appropriate structure.

Predictive modeling requires:

• Target variable:  The target variable column will consist of the customer’s choice. Examples include:
  • 0 or 1 for buy/don’t buy (binary choice) 
  • A, B, or C (e.g., Big Spender/Small Spender/Inactive) for analysis of multiple-choice alternatives
  • A vector with 0 and 1 for analysis of multiple-choice alternatives
  • A number between -infinity and + infinity (e.g., 280, -175, 1,024) for continuous data
• Predictive variables: A column of data for each independent variable specified in the study. Independent variables can take on discrete values if they are appropriately specified using dummy-variable coding, or a consistent set of text (e.g., Rural, Urban).

Optional data requirements:

• Out-of-sample data: If you selected Out-of-sample predictions when setting up your template, you will see a data block for the Out-of-sample data.

Below are examples of each type of data set:

Choice between 2 alternatives (Active = 1, Not active = 0)

Choice between multiple alternatives, one line per choice (Segment = Big Spender/Small spender/Inactive)

Choice between multiple alternatives, one line per alternative (0/0/1)

Note: Each customer can select from a subset of the total set of choice alternatives (here, Brands), in which case, the number of rows may vary across customers.

Continuous choice data (X)

Discrete-continuous choice data (0/X)