How can I handle relational data in machine learning classification? -

i trying classify incidents true positive , false positives using machine learning.

i have dataset of incidents every column describes attribute of incident. , there's list of alerts associated each incident. list of alerts can contain 0-10 alerts every alert row having details of alert. i.e there 1 many relationship between incident , alerts.

i have experience in classifying simple datasets have set of columns every row not sure how handle relational data this.

i using scikit-learn this.

as far understand data looks this:

incident table: id | i_attr0 | alerts 0  | foo     | [alert0, alert1] ...  alert table: id     | a_attr0 alert0 | bar alert1 | baz ... 

if case denormalize table like:

incident-alert table: id | i_attr0 | alert0 | alert0_a_attr0 | alert1 | alert1_a_attr0 | etc.. 0  | foo     | true   | bar            | true   | baz            |  

and work there.