Exam SRM Study Plan: Balancing Theory and Application

Understanding the Exam Structure

Exam SRM (Statistics for Risk Modeling) tests your ability to apply statistical learning methods to actuarial problems. The exam covers linear models, GLMs, time series, tree-based methods, clustering, principal components, and model assessment. Questions test both conceptual understanding (interpreting output, choosing appropriate methods) and computational skills (fitting models, computing metrics). The exam is multiple choice with a mix of calculation and concept questions.

Based on the syllabus weighting, allocate approximately 25-30% of your study time to linear models and GLMs, 20-25% to machine learning methods, 15-20% to time series, and 20-25% to model assessment and selection.

Study Strategy: Weeks 1-6

Weeks 1-2: Linear regression, multiple regression, residual analysis, and variable selection. Focus on interpreting output, understanding assumptions, and diagnostics. Practice computing R-squared, adjusted R-squared, and performing F-tests. Weeks 3-4: GLMs (logistic, Poisson, gamma regression), maximum likelihood estimation, and Bayesian fundamentals. Understand link functions, deviance, and AIC. Weeks 5-6: Time series (AR, MA, ARMA, ARIMA, exponential smoothing). Practice identifying models from ACF/PACF plots and computing forecasts.

Study Strategy: Weeks 7-10

Weeks 7-8: Tree-based methods, random forests, gradient boosting, SVMs, neural networks, and clustering. Focus on understanding algorithms conceptually, their strengths and weaknesses, and hyperparameter tuning. Weeks 9-10: Cross-validation, bias-variance tradeoff, regularization, model evaluation metrics, and review. Take practice exams under timed conditions. For each method, ensure you can: (1) explain when to use it, (2) interpret the output, (3) identify potential problems, and (4) compare it to alternatives. Practice exams from the SOA are the most important study resource for calibrating to the exam's difficulty and style.