Fair Active Learning: Solving the Labeling Problem in Insurance

TL;DR

This paper introduces a novel fair active learning method for insurance models that balances predictive accuracy and fairness, addressing bias and reducing labeling efforts in sensitive data contexts.

Contribution

It proposes an innovative sampling approach that combines informativeness and fairness, improving model fairness without sacrificing performance.

Findings

Effective balancing of fairness and accuracy demonstrated on insurance datasets

Sampling method reduces bias and enhances fairness in model inferences

Numerical experiments confirm improved fairness metrics

Abstract

This paper addresses significant obstacles that arise from the widespread use of machine learning models in the insurance industry, with a specific focus on promoting fairness. The initial challenge lies in effectively leveraging unlabeled data in insurance while reducing the labeling effort and emphasizing data relevance through active learning techniques. The paper explores various active learning sampling methodologies and evaluates their impact on both synthetic and real insurance datasets. This analysis highlights the difficulty of achieving fair model inferences, as machine learning models may replicate biases and discrimination found in the underlying data. To tackle these interconnected challenges, the paper introduces an innovative fair active learning method. The proposed approach samples informative and fair instances, achieving a good balance between model predictive performance and fairness, as confirmed by numerical experiments on insurance datasets.