Reasoning-Enhanced Rare-Event Prediction with Balanced Outcome Correction

TL;DR

This paper introduces LPCORP, a two-stage framework that enhances rare-event prediction by combining reasoning-based predictions with confidence-driven outcome correction, significantly improving performance in highly imbalanced datasets.

Contribution

The novel LPCORP framework effectively balances rare-event datasets without resampling, improving precision and operational cost savings in critical domains.

Findings

Substantially improved precision in low-prevalence datasets

Over 50% reduction in costs with predictive preventive interventions

Transforms imbalanced data into well-balanced predictions

Abstract

Rare-event prediction is critical in domains such as healthcare, finance, reliability engineering, customer support, aviation safety, where positive outcomes are infrequent yet potentially catastrophic. Extreme class imbalance biases conventional models toward majority-class predictions, limiting recall, calibration, and operational usefulness. We propose LPCORP (Low-Prevalence CORrector for Prediction)*, a two-stage framework that combines reasoningenhanced prediction with confidence-based outcome correction. A reasoning model first produces enriched predictions from narrative inputs, after which a lightweight logistic-regression classifier evaluates and selectively corrects these outputs to mitigate prevalence-driven bias. We evaluate LPCORP on real-world datasets from medical and consumer service domains. The results show that this method transforms a highly imbalanced setting into a well-balanced one while preserving the original number of samples and without applying any resampling strategies. Test-set evaluation demonstrates substantially improved performance, particularly in precision, which is a known weakness in low-prevalence data. We further provide a costreduction analysis comparing the expenses associated with rare-event damage control without preventive measures to those incurred when low-cost, prediction-based preventive interventions are applied that showed more than 50% reduction in some cases. * Patent pending: U.S. Provisional 63/933,518, filed 8 December 2025.