-
Fréchet regression with implicit denoising and multicollinearity reduction
Authors:
Dou El Kefel Mansouri,
Seif-Eddine Benkabou,
Khalid Benabdeslem
Abstract:
Fréchet regression extends linear regression to model complex responses
in metric spaces, making it particularly relevant for multi-label regression,
where eachinstance can have multiple associated labels. However, addressing
noise and dependencies among predictors within this framework remains un derexplored. In this paper, we present an extension of the Global Fréchet re gression model tha…
▽ More
Fréchet regression extends linear regression to model complex responses
in metric spaces, making it particularly relevant for multi-label regression,
where eachinstance can have multiple associated labels. However, addressing
noise and dependencies among predictors within this framework remains un derexplored. In this paper, we present an extension of the Global Fréchet re gression model that enables explicit modeling of relationships between input
variables and multiple responses. To address challenges arising from noise
and multicollinearity, we propose a novel framework based on implicit regu larization, which preserves the intrinsic structure of the data while effectively
capturing complex dependencies. Our approach ensures accurate and efficient
modeling without the biases introduced by traditional explicit regularization
methods. Theoretical guarantees are provided, and the performance of the
proposed method is demonstrated through numerical experiments.
△ Less
Submitted 29 March, 2025; v1 submitted 24 December, 2024;
originally announced December 2024.
-
Implicit Regularization for Multi-label Feature Selection
Authors:
Dou El Kefel Mansouri,
Khalid Benabdeslem,
Seif-Eddine Benkabou
Abstract:
In this paper, we address the problem of feature selection in the context of multi-label learning, by using a new estimator based on implicit regularization and label embedding. Unlike the sparse feature selection methods that use a penalized estimator with explicit regularization terms such as $l_{2,1}$-norm, MCP or SCAD, we propose a simple alternative method via Hadamard product parameterizatio…
▽ More
In this paper, we address the problem of feature selection in the context of multi-label learning, by using a new estimator based on implicit regularization and label embedding. Unlike the sparse feature selection methods that use a penalized estimator with explicit regularization terms such as $l_{2,1}$-norm, MCP or SCAD, we propose a simple alternative method via Hadamard product parameterization. In order to guide the feature selection process, a latent semantic of multi-label information method is adopted, as a label embedding. Experimental results on some known benchmark datasets suggest that the proposed estimator suffers much less from extra bias, and may lead to benign overfitting.
△ Less
Submitted 18 November, 2024;
originally announced November 2024.
-
A study of risk management in cloud computing bank
Authors:
Dou El Kefel Mansouri,
Mohamed Benyettou
Abstract:
Cloud computing apparently helps in reducing costs and providing the scheduling optimal level. In practice however it may confront the problem of unavailability of resources. Taking into consideration the cloud computing bank with its somehow commercial nature, the resources unavailability, such as liquidity risk, remains. In this paper, an attempt to show through a solution so far applied in econ…
▽ More
Cloud computing apparently helps in reducing costs and providing the scheduling optimal level. In practice however it may confront the problem of unavailability of resources. Taking into consideration the cloud computing bank with its somehow commercial nature, the resources unavailability, such as liquidity risk, remains. In this paper, an attempt to show through a solution so far applied in economy, how would it be possible to predict such a liquidity risk in cloud computing bank. The proposed solution can especially be adapted to stock management. To reduce the risk we will also make use of a method inspired from physics based on the fluids mechanics; it is an application of Bernoulli's theorem called Torricelli. The resource bank will be considered as a reservoir of liquid, and the availability of resources then will depend on the liquid flow velocity and the replacement.
△ Less
Submitted 10 March, 2014;
originally announced March 2014.
