Setting up a Contextual Bandit

Once you've logged in and chosen to create new experiment from the landing page (see this guide), here's how to set up your CMAB.

Fill out the experiment details

Enter the details for the experiment. Both Experiment Name field and Description are mandatory. Select Contextual Bandit as the experiment type.

Click on the Next button to proceed.

Choose your prior and outcome type

Choose the outcome type: binary or real-valued (the prior is fixed to a normal distribution).

Click on the Next button to proceed.

Configure user context

Enter the user information that you want to track for the experiment. Each context corresponds to one "feature" of the user. The Name and Description fields are mandatory, and the context value can either be real-valued (e.g. age) or binary (e.g. whether or not a user has children).

Then click Next to proceed.

Create experiment arms

Click on the + Add Arm button to add arms to the experiment. You can add as many arms as you like but must have a minimum of 2 arms.

Enter details for each arm. The Name and Description fields are both mandatory.

The priors are defined through the Mean and Standard Deviation parameters of the Gaussian distributions: the default values are 0 and 1 respectively.

Click on the Next button to proceed.

Once you've set up notfications and created the experiment, you can now run the experiment with your users.