FSU mathematician Alec Kercheval has developed new methods to reduce sampling errors in high-dimensional financial data sets. The approach may also have implications for the accuracy and reliability of research findings in fields such as genomics, neuroscience, and image analysis.

Kercheval's research aims to address the challenge of estimating population parameters from high-dimensional data by reducing sampling errors using a combination of machine learning and statistics. Kercheval uses a method called "randomized low-rank approximation" to analyze the data and shrink the sampling error, leading to more accurate estimations of population parameters.

• Alec Kercheval