1
-
5
of
5
results (0.38 seconds)
Sort By:
-
Adapting Credibility Blended Rates to Encompass Machine Learning Predictions
Machine Learning Predictions Actuarial concepts of credibility can be extended to encompass additional ... information. For example, historical claim costs and risk scores both leverage diagnosis code information ...- Date: Apr 2021
- Competency: Technical Skills & Analytical Problem Solving
- Publication Name: Predictive Analytics and Futurism Newsletter
- Topics: Predictive Analytics>Modeling techniques
-
Modeling Extreme Events
A significant fraction of insurance risk is characterized by the tail of the claims distribution, where ... Ignoring the tails can lead to underpricing of tail risk and mispricing. Estimation of the tails can ...- Authors: Robert Jason Reed (Jason)
- Date: Jun 2021
- Competency: Technical Skills & Analytical Problem Solving
- Publication Name: Predictive Analytics and Futurism Newsletter
- Topics: Predictive Analytics>Modeling techniques
-
Exploiting While Exploring: An Introduction to Bayesian Optimization
Introduction to Bayesian Optimization An introduction to the methods and parameters for Bayesian Hyperparameter ... start with parametric statistical models. The classic example is the ordinary least squares (OLS) method to ...- Authors: Nicholas Hanewinckel
- Date: Jun 2021
- Competency: Results-Oriented Solutions; Technical Skills & Analytical Problem Solving
- Publication Name: Predictive Analytics and Futurism Newsletter
- Topics: Life Insurance; Predictive Analytics>Modeling techniques; Predictive Analytics>Programming languages & tools
-
Surrogate Models: A Comfortable Middle Ground?
this article we examine the benefits and drawbacks of using a global surrogate of a black-box model. Are ... models[1] are used for the purpose of model explainability, approximating the predictions from a complex ...- Authors: Harrison Jones
- Date: Aug 2021
- Competency: Technical Skills & Analytical Problem Solving
- Publication Name: Predictive Analytics and Futurism Newsletter
- Topics: Modeling & Statistical Methods>Regression analysis; Predictive Analytics>Modeling techniques
-
GET BETTER ACQUAINTED WITH YOUR KNOWN UNKNOWNS
YOUR KNOWN UNKNOWNS Uncertainty of a predictive model is a fact of life that many insurers could be ... framework for assessing it. The article provides a framework to understand the prediction uncertainty, which ...- Authors: Dan Kim, Boyang Meng
- Date: Apr 2020
- Competency: Results-Oriented Solutions
- Publication Name: Predictive Analytics and Futurism Newsletter
- Topics: Experience Studies & Data>Mortality; Predictive Analytics>Modeling techniques; Predictive Analytics>Quality control & model governance