Support Vector Machine Variable Selection with Lasso and Fused Lasso
| المصدر: | مجلة القادسية للعلوم الإدارية والاقتصادية |
|---|---|
| الناشر: | جامعة القادسية - كلية الادارة والاقتصاد |
| المؤلف الرئيسي: | Dikheel, Tahir R. (Author) |
| مؤلفين آخرين: | Aswad, Zahraa K. (Co-Author) |
| المجلد/العدد: | مج24, ع4 |
| محكمة: | نعم |
| الدولة: |
العراق |
| التاريخ الميلادي: |
2022
|
| الصفحات: | 274 - 286 |
| ISSN: |
1816-9171 |
| رقم MD: | 1345616 |
| نوع المحتوى: | بحوث ومقالات |
| اللغة: | الإنجليزية |
| قواعد المعلومات: | EcoLink |
| مواضيع: | |
| كلمات المؤلف المفتاحية: |
Support Vector Machine | Variable Selection | Lasso | Fused Lasso
|
| رابط المحتوى: |
الناشر لهذه المادة لم يسمح بإتاحتها. |
| المستخلص: |
The support vector machine (SVM) is a highly accurate and adaptable binary classification technique. It has had considerable success, but its performance can suffer if too many covariates are included. In this article, we used lasso penalizes least squares regression by adding the absolute values of the coefficients (L1 - norm). This penalty's structure promotes sparse solutions (with many variables coefficients equal to 0). We propose the fused lasso as a generalization designed for situations like this. Both the simulation study and colon cancer data example show that proposed methods outperform the other existing methods. |
|---|---|
| ISSN: |
1816-9171 |
عناصر مشابهة
-
Support Vector Machine Variable Selection with Lasso and Group Lasso
بواسطة: Dikheel, Tahir R. منشور: (2022) -
Simulation Study for the Reciprocal Lasso Right Censored Response Variable
بواسطة: Mohammed, Fatima Kadhim منشور: (2022) -
A Comparison between Linear Regression-Lasso Quantile Regression Methods in Selecting Best Subset Variables
بواسطة: Ahmed, Neveen Sayed منشور: (2014) -
Simulation Study for the Bayesian Reciprocal Adaptive Lasso Quantile Regression
بواسطة: Lafta, Zahraa Awaid منشور: (2022) -
Support Vector Machines and Fuzzy Nonlinear Regression for Intelligent Identification of urban VANET Constraints
بواسطة: Bechir, Alaya منشور: (2023)
