Abstract
In this work we focus on the problem of local inference for functional data. We describe a unified framework for testing hypotheses on functional data in a local perspective. The result of the testing procedures within the unified framework is an adjusted p-value function that can be used to select the areas of the domain responsible for the rejection of the null hypothesis. We discuss how different state of the art inferential procedures fall within the framework, and briefly describe a novel testing procedure with sound theoretical properties.
Lingua originale | English |
---|---|
Titolo della pubblicazione ospite | Statistics for Smart Applications Book of short papers SIS 2019 |
Pagine | 623-628 |
Numero di pagine | 6 |
Stato di pubblicazione | Pubblicato - 2019 |
Evento | smart statistics for smart applications SIS 2019 - Milano Durata: 19 giu 2019 → 21 giu 2019 |
Convegno
Convegno | smart statistics for smart applications SIS 2019 |
---|---|
Città | Milano |
Periodo | 19/6/19 → 21/6/19 |
Keywords
- nonparametric inference, functional data analysis, family-wise error rate