Transfer Learning of High-Dimensional Stochastic Frontier Model via Elastic Net
In this paper, the high-dimensional stochastic frontier model problem is explored via Elastic Net under the transfer learning framework. When the target data is limited, transfer learning improves the accuracy of model estimation and prediction by transferring the source data. When the transferable...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-06-01
|
| Series: | Axioms |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2075-1680/14/7/507 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | In this paper, the high-dimensional stochastic frontier model problem is explored via Elastic Net under the transfer learning framework. When the target data is limited, transfer learning improves the accuracy of model estimation and prediction by transferring the source data. When the transferable source is known, a transfer learning algorithm for a high-dimensional stochastic frontier model is proposed based on Elastic Net. In addition, based on the prior knowledge of the parameters, this paper introduces linear constraints to improve the estimation accuracy in transfer learning. When the transferable source is unknown, this paper designs a corresponding algorithm to detect the transferable source. Finally, the effectiveness of the method is proved by simulation experiments and actual cases. |
|---|---|
| ISSN: | 2075-1680 |