Lasso and ridge regression methods and their application in GDP deflator estimation analysis

Ziyu Song

doi:10.1117/12.2628045

22 April 2022 Lasso and ridge regression methods and their application in GDP deflator estimation analysis

Ziyu Song

Proceedings Volume 12163, International Conference on Statistics, Applied Mathematics, and Computing Science (CSAMCS 2021); 1216337 (2022) https://doi.org/10.1117/12.2628045
Event: International Conference on Statistics, Applied Mathematics, and Computing Science (CSAMCS 2021), 2021, Nanjing, China

Abstract

Ridge regression and Lasso refer to two types of regression methods that make up some defects existing in OLS, like OLS regression estimator does not uniquely exit when 𝑥𝑥^T is singular, by regularization. This paper reviews and compares the ideas of OLS, ridge regression and Lasso, and sums their main development and previous scholars’ study results. This paper presents that ridge regression adds the 𝑙₂ penalty term as a constraint, and Lasso adds the 𝑙₁ penalty so that it has a function of select variables, some weakly relative variables’ correlated coefficient is reduced to zero directly. By comparing to OLS, ridge regression and Lasso are biased so the choice of these three requires people to consider the specific situation of use. Facing the problem of short-term hysteresis in the calculation of GDP deflator, this paper uses ridge regression and Lasso in GDP deflator analysis by the regression model of GDP deflator with 14 variables, and compares their results with OLS’s. Meanwhile, post Lasso estimator is included to solve the problem of bias at the end, and the conclusion shows the core inflation is the most economically significant variable to GDP deflator in current period estimation.

Citation Download Citation

Ziyu Song "Lasso and ridge regression methods and their application in GDP deflator estimation analysis", Proc. SPIE 12163, International Conference on Statistics, Applied Mathematics, and Computing Science (CSAMCS 2021), 1216337 (22 April 2022); https://doi.org/10.1117/12.2628045

ACCESS THE FULL ARTICLE

INSTITUTIONAL
Select your institution to access the SPIE Digital Library.

SELECT YOUR INSTITUTION

PERSONAL
Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.

PERSONAL SIGN IN

No SPIE Account? Create one

PURCHASE THIS CONTENT

SUBSCRIBE TO DIGITAL LIBRARY

50 downloads per 1-year subscription

Members: $195

Non-members: $335 ADD TO CART

25 downloads per 1 - year subscription

Members: $145

Non-members: $250 ADD TO CART

PURCHASE SINGLE ARTICLE

Includes PDF, HTML & Video, when available