• Model explainability
We prioritized transparency by offering comprehensive visibility into the inner workings of our ML models. By providing in-depth insights into crucial indicators and drivers, the system allows users to validate, investigate, and take specific actions based on observed patterns. This functionality goes beyond mere risk scores, enabling users to understand the underlying factors contributing to piracy.
Design thinking: Building upon the principles of transparency, this module places a strong emphasis on engaging human users in decision-making processes and tailoring interventions. It encompasses two critical components that facilitate this approach:
• Risk-based segmentation
By segmenting users based on their risk levels, we tailored the actions and interventions for each segment. This targeted approach addressed the challenge of personalizing interventions at the user level, while also enabling broader cohort-level interventions, which are highly effective.
© 2023 Fractal Analytics Inc. All rights reserved
06
Made with FlippingBook - PDF hosting