Abstract—The evaluation index system for China’s financial
risk (CFR) and its single-index assessment criteria were used to
generate sufficient samples. The projection pursuit clustering
(PPC) model with a new objective function is applied to
evaluate the CFR during 1994 to 2012. Four rules for
establishing valid PPC model are put forward in this paper, and
the real optimal projection vector coefficients as well as others
parameters are obtained by applying artificial bee colony (ABC)
algorithm. Positive research shows that PPC model is suited to
evaluate CFR, and a valid PPC model as well as six sub-system
models for six aspects of CFR is established. The CFR level
during 1994 to 2012 generally lies on Proper-safety level except
that in 2008 belongs to Risk level. In terms of the six aspects of
CFR, the bank risk is the greatest and the second is the risk of
economic growth.
Index Terms—Artificial bee colony (ABC) algorithm,
China’s financial risk, evaluation index system, projection
pursuit clustering (PPC) model.
Jitong Lou is with the Shanghai University, Shanghai, 200444, P R China
(e-mail: jitonglou@gmail.com).
Wengao Lou and Xiurong Yu are with the Shanghai Business School,
Shanghai, 200235, P R China, corresponding author
(e-mail:wlou64@126.com, wglou@sbs.edu.cn, yuxr11@163.com).
[PDF]
Cite:Jitong Lou, Wengao Lou, and Xiurong Yu, "Evaluation of China’s Financial Risk Using Projection
Pursuit Clustering Model and Its Positive Research," Journal of Economics, Business and Management vol. 2, no. 2, pp. 139-146, 2014.