Optimal Rates for Vector-Valued Spectral Regularization Learning Algorithms
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866913359897559040 |
|---|---|
| author | Meunier, Dimitri Shen, Zikai Mollenhauer, Mattes Gretton, Arthur Li, Zhu |
| author_facet | Meunier, Dimitri Shen, Zikai Mollenhauer, Mattes Gretton, Arthur Li, Zhu |
| contents | We study theoretical properties of a broad class of regularized algorithms with vector-valued output. These spectral algorithms include kernel ridge regression, kernel principal component regression, various implementations of gradient descent and many more. Our contributions are twofold. First, we rigorously confirm the so-called saturation effect for ridge regression with vector-valued output by deriving a novel lower bound on learning rates; this bound is shown to be suboptimal when the smoothness of the regression function exceeds a certain level. Second, we present the upper bound for the finite sample risk general vector-valued spectral algorithms, applicable to both well-specified and misspecified scenarios (where the true regression function lies outside of the hypothesis space) which is minimax optimal in various regimes. All of our results explicitly allow the case of infinite-dimensional output variables, proving consistency of recent practical applications. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2405_14778 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Optimal Rates for Vector-Valued Spectral Regularization Learning Algorithms Meunier, Dimitri Shen, Zikai Mollenhauer, Mattes Gretton, Arthur Li, Zhu Machine Learning We study theoretical properties of a broad class of regularized algorithms with vector-valued output. These spectral algorithms include kernel ridge regression, kernel principal component regression, various implementations of gradient descent and many more. Our contributions are twofold. First, we rigorously confirm the so-called saturation effect for ridge regression with vector-valued output by deriving a novel lower bound on learning rates; this bound is shown to be suboptimal when the smoothness of the regression function exceeds a certain level. Second, we present the upper bound for the finite sample risk general vector-valued spectral algorithms, applicable to both well-specified and misspecified scenarios (where the true regression function lies outside of the hypothesis space) which is minimax optimal in various regimes. All of our results explicitly allow the case of infinite-dimensional output variables, proving consistency of recent practical applications. |
| title | Optimal Rates for Vector-Valued Spectral Regularization Learning Algorithms |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2405.14778 |