Zeyi Wang

Eliminating accidental deviations to minimize generalization error and maximize replicability: applications in connectomics and genomics

This paper introduces a new statistic called discriminability to improve replicability by distinguishing accidental from systematic deviations in data, showing its effectiveness in both simulated and real datasets.

On statistical tests of functional connectome fingerprinting

The paper investigates statistical tests for functional connectome fingerprinting, showing a Poisson(1) approximation applies and highlighting sensitivity issues in permutation tests using fMRI datasets.