Eigencorneas: application of principal component analysis to corneal topography

Rodríguez, Pablo; Navarro, Rafael; Rozema, J.

doi:10.1111/opo.12155

Cited by 11 publications

(13 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…By this method, they increased the weight of the direction of the component with large estimation error and reduced the influence of other directions at the same time. Rodríguez et al [13] yielded a strong reduction of dimensionality of elevation topography data, to only 19 independent parameters combining Zernike fit and PCA and got the minimum number of orthonormal basis functions. Goudarzi et al [14] proposed a noble quantitative structure-property relationship technique on the basis of the random forest.…”

Section: Related Workmentioning

confidence: 99%

Clustering Analysis on the Introduction of Talents in Colleges

Fang¹

2018

IJDST

View full text Add to dashboard Cite

Abstract:With the development of economy and technology, introducing and training talents have become the key driving force in the world which can enhance the competitive strength of the whole countries. Therefore, the strategies of strengthening the universities and colleges with more talented people and making efforts to implement the construction of "Double top" are put forward in the same time. Methods of clustering analysis have been widely used in the actual researches. In this study, an effective clustering analysis model by comparing the clustering analysis under different dimensionality reduction methods is established. Firstly, preprocess the data about talent introduction which is collected from Zhejiang University of Finance and Economics, and use Principal Component Analysis (PCA), Weighted Principal Component Analysis (Weighted-PCA) and Random Forest (RF) to reduce the dimensions of the data. Next, use K-means clustering algorithm and K-medoids clustering algorithm to cluster the preprocessed data. The classification results indicate that the K-medoids algorithm with Weighted-PCA is superior to other clustering methods in this illustrative case. In addition, the experiment divides talents into high-end talents and mid-end talents. By looking into the analysis of the characteristics of the clustering results, some targeted advices on the talents introduction in colleges can be provided.

show abstract

Section: Related Workmentioning

confidence: 99%

Clustering Analysis on the Introduction of Talents in Colleges

Fang¹

2018

IJDST

View full text Add to dashboard Cite

show abstract

“…This lead to a set of 97 biometric parameters (see Table 1 for an overview), which was reduced down to 18 by processing the corneal elevation Zernike coefficients using Principal Component Analysis ("eigencorneas"). 11 Next, these parameters were fitted with a linear combination of two multivariate Gaussian functions using Expectation-Maximization, 12 from which it is possible to generate an unlimited number of random biometry sets with the same distributions as the original data. 10,13 Finally, these biometry sets (or Syn-tEyes) were filled in into the Navarro eye model 7 to account for missing parameters, such as e.g.…”

Section: Measurements and Modellingmentioning

confidence: 99%

SyntEyes KTC: higher order statistical eye model for developing keratoconus

Rozema

Rodríguez

Hidalgo

et al. 2017

Ophthalmic Physiologic Optic

Self Cite

View full text Add to dashboard Cite

show abstract

“…As the large number of parameters involved could cause dimensionality problems during the Gaussian modeling, we used principal component analysis to compress the number of dimensions from 91 corneal parameters to 12 eigenvectors (eigencorneas [ECs]) while retaining 99.5% of the original variability. 15 The remaining 0.5% of variability lies mostly in the higher-order aberrations and is of no real consequence to the model (Supplementary Material A). These eigencorneas are used during the stochastic process and converted back to Zernike coefficients and CCT afterward.…”

Section: Modeling Of the Eyementioning

confidence: 99%

“…The purpose of the current work is therefore to improve the accuracy of the earlier model by including a Zernike description of the corneal shape. Furthermore, a number of technical improvements will be introduced to enhance its performance, such as principal component analysis to reduce the number of parameters, 15 more accurate methods to estimate the crystalline lens shape 16 and power, 17 and a linear combination of multivariate Gaussians 5 instead of refractive filtering. After verification of the model, examples will be given of possible applications.…”

mentioning

confidence: 99%