August 31, 2021
The CSPA Exam 3 Predictive Modeling – Methods and Techniques committee has revised the syllabus effective with the upcoming November 4 exam sitting. The committee did not add any material but rather transferred a sizeable amount of material from the exam that they felt was better suited to continuing education.
The changes are as follows:
- Section A4 – Hierarchical Models, including Linear Mixed Models, and Bühlmann Credibility – has been removed altogether
- Section B8 – Principal Components Analysis and Unsupervised Learning – has been shortened to introduce Unsupervised Learning and cover clustering methods (but eliminate principal components).
- Sections B9 and B10 – Application Specific Methods: Association Models and Application Specific Methods: Fraud Detection have both been removed
- Section C – including C1, C2, and C3 – Causal Inference – has been removed in totality except for one reading which has been relocated to a new section A4 – Causal Inference from Observational Data
While the syllabus has been condensed, the exam will continue to be 4-hours in length.
Continuing education opportunities covering the removed material are under development.
Please contact iCAS Director Amy Brener with any questions.
