A0665
Title: Inference on functional data through e-values
Authors: Alessia Pini - Universita Cattolica del Sacro Cuore (Italy) [presenting]
Abstract: A very recent area of statistical inference proposes to replace p-values with e-values for testing hypotheses on univariate or multivariate data. Namely, an e-value (also known as a betting score) is the value taken by a random variable whose expected value is equal to one under the null hypothesis. The extension of e-values is proposed to functional data. In detail, it starts by defining a point-wise e-value function for performing inference on each point of the domain of the functions, separately. Then, it is shown that the average of the point-wise e-value function is a valid e-value for performing global inference on functional data. Finally, it is discussed how to adjust the point-wise e-value function for performing local inference.
