Title
Fast assessment of freshness of white shrimp via a modified unsupervised discriminant projection technique
Abstract
Currently, fish industry lacks the ability to quantitate the degree of freshness of white shrimp. In order to find a simple method for determining shrimp quality, the use of electronic nose was investigated, but the traditional classification algorithms are intrinsically not suitable for electronic nose data which has non-linear manifold structures. Hence a modified unsupervised discriminant projection (MUDP) coupled with sample label information was proposed. MUDP can keep the local and global structure and can take advantage of the important of label information, then get geometric structure optimal linear projection. Experimental results indicate that the proposed classification algorithm is much better than some traditional algorithms such as PCA, LDA, etc.
Year
Venue
Keywords
2015
International Conference on Control Automation and Information Sciences
electronic tongue,manifold learning,KPCA,white shrimp
Field
DocType
ISSN
Electronic nose,Sample Label,Global structure,Pattern recognition,Discriminant,Projection (linear algebra),Artificial intelligence,Engineering,Statistical classification,Principal component analysis,Shrimp
Conference
2475-7896
Citations 
PageRank 
References 
0
0.34
2
Authors
4
Name
Order
Citations
PageRank
Peiyi Zhu121.39
Jie Du200.34
Chen-Sheng Chen300.34
Xiaoyun Gu400.34