PredicTri - Content Library

📖 Content Library

Explore our comprehensive collection of articles, case studies, and insights on AI/ML in actuarial practice

Dive deep into the latest developments in Bayesian machine learning, model validation, and next-generation triangle modeling techniques.

1️⃣ AI/ML in Actuarial Practice

Why ML for Actuaries?

What is Machine Learning?

Next-Gen Triangle Modeling: Stochastic, Bayesian, Explainable

AI Adoption Barriers & Explainability

Insurance is Conservative: Why Our Tool is a Second Opinion, Not a Replacement

Disagreeing with the Machine: What Comes Next?

ML Workflows & Computational Costs

Why Don’t We Get Instant Results from Machine Learning Models?

Generative Models & Risk Components

Teaching your Models - Augmenting vs Understanding

2️⃣ Actuarial Modeling Insights

Model Explainability & Transparency

Determining the Influence of Factors in Complex/Non-Linear Actuarial Models? (Shapley values explained)

Joint Paid & Incurred Modeling

How to Make Paid and Incurred Models Agree

When Trends Don’t Tell the Whole Story (Incurred vs Paid trends and joint models)

Joint Modeling in Health Insurance: A More Informative Projection

Model Validation & Cross-Validation

How Do You Pick a Winning Model?

Model Validation

When Traditional Actuarial Practice is Actually Cross-Validation

Multi-Fold Validation in Reserving Models

Cross-Validation & Overfitting Prevention

Comparing Predictions with Actual Results: Energy Score Explained

Understanding Percentiles & Predictive Distributions

Why Uniform Distribution of Percentiles Matters

Model Percentiles of Observed Ultimate Loss Ratios: Sparse Data Case Study

How to Tell If Your Predictive Distribution Is Right

Case Study: Long-Tail Liability Portfolio

Why Standard Models Miss the Extremes: A Case for Better Residuals

Non-Cumulative Models vs Traditional

A Good Start Means a Good Finish — Or Does It?

Beyond Tradition: Testing Non-Cumulative Models in Claims Development

When the Weather is Too Good: Rethinking Assumptions with ML

Good or Bad Start? How Initial Development Affects Reserve Risk Estimation

Sparse Data & Difficult Portfolios

Model Percentiles of Observed Ultimate Loss Ratios: Sparse Data Case Study

Modeling a Buses Bodily Injury Portfolio

Case Study: Long-Tail Liability Portfolio

Bayesian Modeling

Why Bayesian Models Shine on Sparse Triangles

Unified Modeling & Workflow Efficiency

One Model, All the Answers

Case Study Notebooks

Checking Outcome Distributions Against Model Predictions

3️⃣ From the Community

Embedding Actuarial Insight Into Machine Learning

Don’t Worry About Runtime — Worry About ROI

Reserving Reinvented: Bringing Modern Risk Modeling to Nepal with PredicTri

The Value of Explainable Models in Specialty Lines

The Evolving Role of AI/ML in Actuarial Reserving

Thoughts on Actuarial Innovation: Method, Implementation & Incentives

PredicTri: A Game-Changer in Actuarial Reserving

Words That Resonate