Hi @oumera64,

only scalar properties can be used as a predictor in the ADM model (in fact all the models used in Pega or outside Pega). 

Here’s how you can approach it to transform the PageList predictors to scalar values.  

1. Feature Engineering:

   • Create a feature (predictor) that represents the list.
     • For instance, if you have a list of product ratings, you could compute the average rating or the total count of ratings as a single feature.
     • Similarly, if you have a list of transaction amounts, you might calculate the sum or average of those amounts.

2. Flatten Nested Lists:

   • If your list contains nested structures (e.g., a list of lists), consider flattening it.
     • For instance, if you have a list of orders, each containing multiple line items, flatten it to a single list of line items.

3. Encoding Categorical Lists:

   • If your list contains categorical data (e.g., product categories), encode it appropriately.
   • You can use techniques like one-hot or label encoding to represent categorical values numerically.

Please do not forget to preprocess your data and handle missing values appropriately. Additionally, test the performance of your model with and without the list-based predictors to assess their impact. 

If you still need assistance, don't hesitate to reach out to me. 

Thanks,

Nanjundan Chinnasamy | Pega Lead Decision Architect | DCS | 1:1 Customer Engagement