The Research of High-Dimensional Big Data Dimension Reduction Strategy

Qing Chao Jiang

doi:10.4028/www.scientific.net/AMM.710.121

Paper Titles

The Measurement of the Fatigue Crack-Tip Displacement and Strain Fields under High Frequency Resonant Loading Applying DIC Method
p.83

Comparative Investigation of an Automated Oceanic Wave Surface Glider Robot Influence on Resistance Prediction Using CFD Method
p.91

Optimal the Multiple Performances in EDM for SKH57 Using Grey Relational Analysis
p.99

Overview of Centrifuge Model Test Development and its Application
p.107

Modeling and Simulation of Solar Radiation
p.113

The Research of High-Dimensional Big Data Dimension Reduction Strategy
p.121

The Research of Generation Algorithm of Frequent Itemsets in High-Dimensional Data
p.127

A Study of Distinguisher Attack on AES-128 and AES-256 Block Ciphers through Model Based Classification Using Neural Network
p.133

Research on Multi-Agent Intelligent Information Retrieval Model
p.139

HomeApplied Mechanics and MaterialsApplied Mechanics and Materials Vol. 710The Research of High-Dimensional Big Data...

The Research of High-Dimensional Big Data Dimension Reduction Strategy

Abstract:

With the increase of data dimension, many low dimensional mining algorithms cannot get satisfactory results. With the increase of data dimension, it can produce a large amount of redundant information; this information will greatly reduce the efficiency of mining, increasing the complexity of the mining algorithm. Feature selection is an efficient way to solve the problem; it can remove a lot of irrelevant and redundant features. In this paper, on the basis of Lars algorithm applying differential evolution thought to the extraction of feature subset, puts forward a new method of feature selection, DE - Lars algorithm. Experiments prove that DE - Lars algorithm enhances the precision of reducing dimension of space, effectively solve the problem of "Curse of Dimensionality ".

You might also be interested in these eBooks

View Preview

Info:

Periodical:

Applied Mechanics and Materials (Volume 710)

Pages:

121-126

DOI:

https://doi.org/10.4028/www.scientific.net/AMM.710.121

Citation:

Cite this paper

Online since:

January 2015

Authors:

Qing Chao Jiang*

Keywords:

Data Mining, DE, High Dimensional Data, Lars, Lasso

Export:

RIS, BibTeX

Price:

Permissions CCC:

Request Permissions

Permissions PLS:

Request Permissions

Сopyright:

Citation:

* - Corresponding Author

References

[1] R. Bellman. Adaptive Control Processes: A Guided Tour [M]. Princeton University Press, Princeton, New Jersey, (1961).

DOI: 10.1002/nav.3800080314

Google Scholar

[2] D.W. Scott. Multivariate Density Estimation [M]. Wiley&Sons, (1992).

Google Scholar

[3] Smialowski R. Frishman D., Kramer S. Pitfalls of supervised feature selection [J]. Bioinformatics, 2010, 26(3): 440-443.

DOI: 10.1093/bioinformatics/btp621

Google Scholar

[4] Fomian G. An extensive empirical study of feature selection metrics for text classification [J]. The Journal of machine learning research, 2003, 3: 1289-1305.

Google Scholar

[5] Rainer Storn, Kenneth Price and Jouni Lampinen. Differential Evolution: A Practical Approach to Global Optimization[M]. Berlin: Springer-Verlag, (2005).

Google Scholar

[6] Rainer Storn. On the usage of differential evolution for function optimization [C]. Proceedings of the North Amer. Fuzzy Inf. Process. Soc, NAFIPS, Berkeley, CA, 1996: 519-523P.

Google Scholar