Class SignalsResponse

  • All Implemented Interfaces:

    
    public final class SignalsResponse
    
                        

    Behavioral feature state for a card or account derived from its transaction history.

    Derived statistical features (averages, standard deviations, z-scores) are computed using Welford's online algorithm over approved transactions. Average fields are null when fewer than 5 approved transactions have been recorded. Standard deviation fields are null when fewer than 30 approved transactions have been recorded.

    3DS fields (three_ds_success_rate, three_ds_success_count, three_ds_total_count) are card-scoped and will be null for account responses.

    Raw fields (seen_countries, seen_mccs, approved_txn_amount_m2, etc.) are included so clients can compute their own transaction-specific derivations, such as checking whether a new transaction's country is in seen_countries to determine is_new_country, or computing a z-score using the raw mean and M2 values.