Estimation and feature selection in high-dimensional mixtures-of-experts models
Doctoral Thesis
Đã lưu trong:
| Tác giả chính: | |
|---|---|
| Định dạng: | Dissertation |
| Ngôn ngữ: | English |
| Được phát hành: |
University of Caen
2023
|
| Những chủ đề: | |
| Truy cập trực tuyến: | https://scholar.dlu.edu.vn/thuvienso/handle/DLU123456789/116213 |
| Các nhãn: |
Thêm thẻ
Không có thẻ, Là người đầu tiên thẻ bản ghi này!
|
| Thư viện lưu trữ: | Thư viện Trường Đại học Đà Lạt |
|---|
| id |
oai:scholar.dlu.edu.vn:DLU123456789-116213 |
|---|---|
| record_format |
dspace |
| spelling |
oai:scholar.dlu.edu.vn:DLU123456789-1162132023-10-05T16:57:25Z Estimation and feature selection in high-dimensional mixtures-of-experts models Huỳnh, Bảo Tuyên Mixture models Mixture of Experts Regularized Estimation Feature Selection Doctoral Thesis This thesis deals with the problem of modeling and estimation of high-dimensional MoE models, towards effective density estimation, prediction and clustering of such heterogeneous and high-dimensional data. We propose new strategies based on regularized maximum-likelihood estimation (MLE) of MoE models to overcome the limitations of standard methods, including MLE estimation with Expectation-Maximization (EM) algorithms, and to simultaneously perform feature selection so that sparse models are encouraged in such a high-dimensional setting. We first introduce a mixture-of-experts' parameter estimation and variable selection methodology, based on ℓ1 (lasso) regularizations and the EM framework, for regression and clustering suited to high-dimensional contexts. Then, we extend the method to regularized mixture of experts models for discrete data, including classification. We develop efficient algorithms to maximize the proposed ℓ1-penalized observed-data log-likelihood function. Our proposed strategies enjoy the efficient monotone maximization of the optimized criterion, and unlike previous approaches, they do not rely on approximations on the penalty functions, avoid matrix inversion, and exploit the efficiency of the coordinate ascent algorithm, particularly within the proximal Newton-based approach. 2023-08-11T06:13:36Z 2023-08-11T06:13:36Z 2019 Dissertation https://scholar.dlu.edu.vn/thuvienso/handle/DLU123456789/116213 en application/pdf University of Caen |
| institution |
Thư viện Trường Đại học Đà Lạt |
| collection |
Thư viện số |
| language |
English |
| topic |
Mixture models Mixture of Experts Regularized Estimation Feature Selection |
| spellingShingle |
Mixture models Mixture of Experts Regularized Estimation Feature Selection Huỳnh, Bảo Tuyên Estimation and feature selection in high-dimensional mixtures-of-experts models |
| description |
Doctoral Thesis |
| format |
Dissertation |
| author |
Huỳnh, Bảo Tuyên |
| author_facet |
Huỳnh, Bảo Tuyên |
| author_sort |
Huỳnh, Bảo Tuyên |
| title |
Estimation and feature selection in high-dimensional mixtures-of-experts models |
| title_short |
Estimation and feature selection in high-dimensional mixtures-of-experts models |
| title_full |
Estimation and feature selection in high-dimensional mixtures-of-experts models |
| title_fullStr |
Estimation and feature selection in high-dimensional mixtures-of-experts models |
| title_full_unstemmed |
Estimation and feature selection in high-dimensional mixtures-of-experts models |
| title_sort |
estimation and feature selection in high-dimensional mixtures-of-experts models |
| publisher |
University of Caen |
| publishDate |
2023 |
| url |
https://scholar.dlu.edu.vn/thuvienso/handle/DLU123456789/116213 |
| _version_ |
1819763599338373120 |