Applying Neural Networks to Study the Mesoscale Variability of Oceanic Boundary Currents

Botelho, Silvia Silva da Costa; Mata, Maurício M.; de, Rodrigo; Almeida, Igor Lorenzato

doi:10.1007/978-3-540-39592-8_100

Cited by 2 publications

(2 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Thus, when the original dataset have many dimensions, several authors opt to filter the data before the NLPCA analysis, like the use of PCA reduction techniques [1], [5]. Using the former approach, the simplification introduced by the use of linear PCA analysis can lead to erroneous outputs or, at least, can produce coarser results.…”

Section: Principal Component Analysis Methods a Principal Compomentioning

confidence: 99%

“…Once Let f: RP -+ r denotes the function modeled by layers 1, 2 and 3, and let s: Rr -* RP denotes the function modeled by layers 3, 4 and 5. Using this notation, the weights in the NLPCA network are determined under the following objective function: n min EII---1l 1=1 (1) , where X' is the output of the network. The relation associated with u and X is now generalized to u = f(X), where f can be any nonlinear function explained by a feed-forward NN mapping from input layer to the bottleneck layer and instead of PCA, X1 -Xe11 is minimized by nonlinear mapping functions, A' = s(u).…”

Section: Principal Component Analysis Methods a Principal Compomentioning

confidence: 99%

See 1 more Smart Citation

A method to extract non-linear principal components of large datasets - an application in skill transfer

Botelho

Bern

Figueiredo

et al. 2005

Proceedings. 2005 IEEE International Joint Conference on Neural Networks, 2005.

View full text Add to dashboard Cite

This article presents a methodology to extract principal components of large datasets, called C-NLPCA (Cascaded nonlinear principal component analysis), and evaluates its use in the extraction of main human movements in image series, aiming for the development of methodologies and techniques for skill transfer from humans to robotic/virtual agents. The C-NLPCA is an original data multivariate analysis method based on the NLPCA (Nonlinear Principal Component Analysis). This method has as main features the capability of taking principal variability components from a large set of data, considering the existence of possible nonlinear relations among them. The proposed method is used to extract principal movements from video sequence of human activities, which can be reconstructed in cybernetic and robotic contexts. Aiming for the validation of the method a human moving hand test is presented, where C-NLPCA is applied and the patterns of the obtained movements are confronted with traditional linear techniques.

show abstract

Section: Principal Component Analysis Methods a Principal Compomentioning

confidence: 99%

Section: Principal Component Analysis Methods a Principal Compomentioning

confidence: 99%