Diversified Learning: Bayesian Control with Multiple Biased Information Sources
42 Pages Posted: 25 Oct 2024
Date Written: October 23, 2024
Abstract
We consider a decision-maker (DM) who can acquire signals from multiple biased information sources to learn about a hidden state before making an earning decision. Unbiased signals are also available but come at a higher acquisition cost. The DM jointly optimizes both learning (information acquisition) and earning decisions to minimize expected loss. This problem is motivated by applications such as medical diagnostics and financial investments, where forecasting and decisions rely on multiple potentially biased information sources. We develop a Bayesian decision framework for a general class of such problems, modeling the multisource learning structure with a hierarchical Bayesian network. We fully solve the model and explicitly characterize the optimal acquisition policy, which diversifies across biased information sources to mitigate the risks associated with bias. We also illustrate the model using disease forecasting and intervention datasets.
Keywords: Multi-source learning, Bayesian network, optimal control, bias and misinformation, diversification strategy, forecasting aggregation
Suggested Citation: Suggested Citation