arXivOther

Uncertainty-Aware Strategies: A Model-Agnostic Framework for Robust Financial Optimization through Subsampling

The paper tackles model uncertainty in quantitative finance by improving the conventional objective with an outer uncertainty measure, suggesting a subsampling strategy to approximate model uncertainty and introducing a modified stochastic gradient descent algorithm for efficient parallelization.