Improving Geopolitical Forecasts with Bayesian Networks

TL;DR

This paper compares Bayesian networks, logistic regression, and recalibration methods for geopolitical forecasting, finding that recalibrated aggregates perform best, but BNs and logistic models offer valuable insights with potential for hybrid approaches.

Contribution

It introduces the application of structure-learned and naive Bayesian networks to geopolitical forecasting and compares their performance to traditional methods using real-world data.

Findings

Recalibrated aggregate achieved highest accuracy (AUC = 0.985).

Bayesian networks performed nearly as well as the aggregate.

Logistic regression models were less accurate, possibly due to linearity violations.

Abstract

This study explores how Bayesian networks (BNs) can improve forecast accuracy compared to logistic regression and recalibration and aggregation methods, using data from the Good Judgment Project. Regularized logistic regression models and a baseline recalibrated aggregate were compared to two types of BNs: structure-learned BNs with arcs between predictors, and naive BNs. Four predictor variables were examined: absolute difference from the aggregate, forecast value, days prior to question close, and mean standardized Brier score. Results indicated the recalibrated aggregate achieved the highest accuracy (AUC = 0.985), followed by both types of BNs, then the logistic regression models. Performance of the BNs was likely harmed by reduced information from the discretization process and violation of the assumption of linearity likely harmed the logistic regression models. Future research should explore hybrid approaches combining BNs with logistic regression, examine additional predictor variables, and account for hierarchical data dependencies.