EURO-Online login
- New to EURO? Create an account
- I forgot my username and/or my password.
- Help with cookies
(important for IE8 users)
3132. Visualizing outliers in functional data with new versions of the epigraph and the hypograph indices
Invited abstract in session WB-27: Unraveling the Black Box: Advances in Model Explainability, stream Mathematical Optimization for XAI.
Wednesday, 10:30-12:00Room: 047 (building: 208)
Authors (first author is the speaker)
| 1. | Belén Pulido Bravo
|
| uc3m-Santander Big Data Institute, Universidad Carlos III de Madrid | |
| 2. | Alba M. Franco-Pereira
|
| Statistics and OR, Universidad Complutense de Madrid | |
| 3. | Rosa Elvira Lillo Rodríguez
|
| Statstics, Universidad Carlos III de Madrid |
Abstract
Outlier detection in functional data presents a critical challenge in contemporary data science research. The inherent complexity of these datasets implies working with an infinite dimensional space. Addressing this challenge demands innovative methodologies that yield robust and interpretable results.
In this study, we introduce novel iterations of the epigraph and hypograph indexes, different from those available in the existent literature. These indices allow to transform the original functional dataset into a multivariate one, enabling the application of available outlier detection methods for multivariate datasets. The inherent visual nature of the epigraph and the hypograph indices facilitates intuitive interpretation of outcomes, enabling users to extract insights into identified outliers.
This methodology is validated through comprehensive experimentation on both simulated and real datasets, demonstrating competitive results compared to existing methodologies available in the literature.
Keywords
- Analytics and Data Science
Status: accepted
Back to the list of papers