Score Models
Recently we updated the large data machine learning models that generate the propensity to pay scores with regulatory changes, new data sources, and enhanced algorithms. We will implement this model upgrade for all our score customers.
The new model tends towards a higher baseline which will have the impact of awarding more recent accounts with higher scores. Our medal algorithms will smooth this bias automatically over time.
Litigation ScoreÂ
A few of our customers use a litigation score meant to predict recovery if a suit is brought against a consumer. Let your consultant know if you’d like to add this score to your solution.Â
Signals
We added a SIG-ID-31-Consumer Opted Out signal for those consumers who opted out of enough public databases to make scoring them impossible for TU. Reach out to your service delivery rep to learn more about Signals.
Charity
In January, we updated the federal poverty guidelines per the Department of Health and Human services. This informs the P, Q, R, S and FPG score values.