ACTEX Study Manual for Exam ATPA - 2nd Edition - Advanced Predictive Analytics Models and Related Issues

Sep 24, 2026

ACTEX Study Manual for Exam ATPA (2nd Edition) provides structured preparation materials covering advanced predictive analytics modeling and advanced data management. It guides learners through generalized additive models, generalized linear mixed models, neural networks, Bayesian regression, and stacking, with hands-on tasks for exploring data, fitting models, tuning hyperparameters, validating results, and making predictions. The manual also addresses model explainability and communication techniques, plus technical, ethical, and legal considerations for responsible data workflows.

LanguageEnglish

CategoryExam

FormatPDF

Pages63

Reading Time2 h 39 min

AccessFree to download

Preface

About Exam ATPA

About this Study Manual

Advanced Predictive Analytics Models and Related Issues

Advanced Predictive Analytics Models

Generalized Additive Models

Generalized Linear Mixed Models

Neural Networks

Bayesian Regression

Stacking

Model Explainability and Communication

Techniques for Interpreting Opaque Models

Techniques for Communicating with the Intended Audience

Advanced Data Management

Advanced Data Management: Technical Issues

Advanced Data Management: Ethical and Legal Issues

ACTEX Learning Learn Today. Lead Tomorrow. Actuarial & Financial Risk Resource Materials Since 1972 Copyright © 2025, ACTEX Learning, a division of ArchiMedia Advantage Inc. No portion of this ACTEX Study Manual may be reproduced or transmitted in any part or by any means without the permission of the publisher. Contents Preface xi P.1 About Exam ATPA .................................. xi P.2 About this Study Manual .............................. xvi I Advanced Predictive Analytics Models and Related Issues 1 1 Advanced Predictive Analytics Models 3 1.1 Generalized Additive Models ............................. 4 1.1.1 Conceptual Foundations ........................... 4 1.1.2 Case Study .................................. 7 TASK 1: Explore and prepare the data ................... 10 TASK 2: Fit and evaluate two GAMs .................... 17 TASK 3: Fine-tune and validate the recommended GAM ......... 26 1.2 Generalized Linear Mixed Models .......................... 31 1.2.1 Conceptual Foundations ........................... 31 1.2.2 Case Study .................................. 38 TASK 1: Explore and prepare the data ................... 39 TASK 2: Fit and evaluate a random intercepts GLMM .......... 43 TASK 3: Fit and evaluate a random slopes GLMM ............ 48 TASK 4: Use GLMMs to make predictions ................. 51 1.3 Neural Networks ................................... 54 1.3.1 Conceptual Foundations ........................... 54 Part 1: Basic Terminology .......................... 55 Part 2: How to Train a Neural Network ................... 62 1.3.2 Case Study 1: Regression .......................... 68 TASK 1: Prepare the data .......................... 69 TASK 2: Build and evaluate a baseline neural network .......... 72 TASK 3: Explore the effects of the mini-batch size and learning rate ... 77 TASK 4: Tune and evaluate the neural network .............. 81 1.3.3 Case Study 2: Classification ......................... 89 TASK 1: Explore the data .......................... 90 TASK 2: Build and evaluate the first network classifier .......... 93 TASK 3: Tune and evaluate the network .................. 100 TASK 4: Tune and evaluate the network .................. 105 1.4 Bayesian Regression ................................. 108 1.4.1 Conceptual Foundations ........................... 108 1.4.2 Case Study .................................. 113 TASK 1: Fit and analyze a Bayesian GLM ................. 113 TASK 2: Compare models and make predictions .............. 120 1.5 Stacking ........................................ 123 1.5.1 Conceptual Foundations ........................... 123 1.5.2 Case Study .................................. 128 TASK 1: Build the stage-0 models ..................... 128 TASK 2: Tune the stage-1 model ...................... 132 2 Model Explainability and Communication 137 2.1 Techniques for Interpreting Opaque Models ..................... 138 2.1.1 Global Method 1: Variable Importance ................... 140 2.1.2 Global Method 2: Partial Dependence ................... 144 2.1.3 Global Method 3: Global Surrogate Models ................ 152 2.1.4 Local Method 1: ICE Plots ......................... 153 2.1.5 Local Method 2: SHAP ........................... 155 2.1.6 Digression: Lift and Gain Charts ...................... 156 2.2 Techniques for Communicating with the Intended Audience ............ 162 2.2.1 Technical Reports ............................... 165 2.2.2 Executive Summaries ............................. 168 II Advanced Data Management 173 3 Advanced Data Management: Technical Issues 175 3.1 Data Pipeline ..................................... 176 3.2 Data Transformation ................................. 178 3.2.1 Data Manipulation on Rows ......................... 180 3.2.2 Data Manipulation on Columns ....................... 188 3.2.3 Grouping and Summarizing Da

FAQs

What modeling topics are covered for Exam ATPA preparation?

The manual covers generalized additive models, generalized linear mixed models, neural networks, Bayesian regression, and stacking, with case studies and structured tasks to fit, evaluate, tune, and validate models.

How does the book approach model explainability?

It presents techniques for interpreting opaque models, including global and local methods such as variable importance, partial dependence, surrogate models, ICE plots, and SHAP, along with related charts like lift and gain.

What data management and compliance areas are included?

It covers technical issues such as data pipelines, transformations, cleaning, and validation, plus ethical principles, regulations and standards of practice, and algorithmic fairness considerations.

Sep 24, 2026

Uploaded by Nguyễn Văn Học