Composite kernel machines on kernel local fisher discriminant space for financial data mining

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper proposes a novel approach to overcome the bottleneck in financial data mining. We construct a composite kernel machine (CKM) on the kernel local fisher discriminant space (KLFDS) to solve three problems in high-dimensional data mining: the curse of dimensionality, data complexity and nonlinearity. CKM exploits multiple data sources with strong capability to identify the relevant ones and their apposite kernel representation. KLFDS is an optimal projection of original data to a low dimensional space which maximizes the margin between data points from different classes at each local area of data manifold. Our new system robustly overcomes the weakness of CKM, it outperforms many traditional classification systems.

Original languageEnglish
Title of host publicationIET International Conference on Information and Communications Technologies, IETICT 2013
Pages58-63
Number of pages6
Volume2013
Edition618 CP
Publication statusPublished - 2013 Dec 1
EventIET International Conference on Information and Communications Technologies, IETICT 2013 - Beijing, China
Duration: 2013 Apr 272013 Apr 29

Other

OtherIET International Conference on Information and Communications Technologies, IETICT 2013
CountryChina
CityBeijing
Period13-04-2713-04-29

All Science Journal Classification (ASJC) codes

  • Electrical and Electronic Engineering

Fingerprint Dive into the research topics of 'Composite kernel machines on kernel local fisher discriminant space for financial data mining'. Together they form a unique fingerprint.

  • Cite this

    Huang, S-C., & Wu, T-K. (2013). Composite kernel machines on kernel local fisher discriminant space for financial data mining. In IET International Conference on Information and Communications Technologies, IETICT 2013 (618 CP ed., Vol. 2013, pp. 58-63)