Work

Exploring Determinants of Inflation in China based on L1-∈-Twin Support Vector Regression

Year: 2013

Type: conference-paper

Abstract: As a novel feature selection approach, L1-norm E-twin support vector regression(L1-E- TSVR)is proposed in this paper to investigate determinants of cost-push inflation in China. Compared with L2-ε-TSVR, our L1-E- TSVR not only can fit function well, but also can do feature ranking. The computational results of inflation forecasts demonstrate that our L1-E- TSVR derives much smaller root mean squared error (RMSE) than the forecasts generated from ordinary least square (OLS) model. Furthermore, the feature selection results indicate that the most significant explanatory factor for the inflation in China is the housing sales price index. Therefore, the housing market do have an important impact on the inflation in China.

Source: Procedia Computer Science

DOI: 10.1016/j.procs.2013.05.066

Authors: Ya-Fen Ye, Hui Cao, Lan Bai, Zhen Wang, Yuan-Hai Shao

Institutions: Zhejiang University of Technology, Jilin University

Language: English

Topic: Monetary Policy and Economic Impact

Subfield: General Economics, Econometrics and Finance

Field: Economics, Econometrics and Finance

Domain: Social Sciences

Open Access status: diamond

Funders: National Natural Science Foundation of China, Jilin University, Department of Education of Zhejiang Province, Natural Science Foundation of Zhejiang Province

Awards: G1857671292, G5797183096, G7277974055, G7394556461