Robust Decisions with Wild Data

From providing mortgages to screening resumes, algorithms make our decisions for us based on vast amounts of data. They have come under scrutiny as they are dangerously gullible in uncertain environments. Robust decisions promise certified protection against uncertainty, following the maxim "Plan for the worst, hope for the best." Robust decisions have proven more trustworthy when based on curated data sets. Unfortunately, robust decisions, too, become gullible when using data scraped online without stringent quality controls. The researchers will develop clever decision algorithms that protect against wild online data and enjoy better typical performance beyond merely the worst-case.