Some remarks on the functional relation between canonical correlation analysis and partial least squares

Malec, Lukáš

doi:10.1080/00949655.2015.1110821

Cited by 9 publications

(5 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In comparison to canonical correlation analysis, the PLS algorithm ranges to define a new coordinate system so that such a pair of coordinate systems is optimal in maximizing the covariance of individual variables of the first data set to LV of the second set (Wold, 1975). Malec (2016) discusses the functional relation for both PLS and canonical analysis. While in the case of PLS-LDA, the distances between LVs are to be maximized, for partial least squares, the covariance relations are in the area of interest.…”

Section: Methodsmentioning

confidence: 99%

Local Food as a Tool of Tourism Development in Regions

Kiráľová

Malec

2021

International Journal of Tourism and Hospitality Management in the Digital Age

Self Cite

View full text Add to dashboard Cite

This study aims to identify the importance of local food for both the demand and supply sides and to show how local food can be bounded with tourism development in the region. The data presented are based on secondary and primary research. Secondary research includes the literature review and content analysis of documents. The qualitative research included a questionnaire survey among guests of the gastronomic establishments and entrepreneurs. Partial least squares variant of linear discriminant analysis (PLS-LDA) and partial least squares (PLS) as an alternative to standard multivariate methods were used to show the gastronomic establishments guests' and entrepreneurs' opinions on local seasonal food and beverages. The opinions are moreover related to the economically driven interest of guests and entrepreneurs. Based on the typical random variable source, data were gathered from three Czech regions covering the scope of this study. The significant disputes between opinions on local food and beverages are directly applicable in practice, including individual items.

show abstract

Section: Methodsmentioning

confidence: 99%

Local Food as a Tool of Tourism Development in Regions

Kiráľová

Malec

2021

International Journal of Tourism and Hospitality Management in the Digital Age

Self Cite

View full text Add to dashboard Cite

show abstract

“…Specifically, in this study, the functional relation examines the analytical (thus also smooth) paths of eigenvalues and corresponding eigenvectors solving the Fisher discriminant analysis (Malec, 2015). Malec (2016) compares canonical correlation analysis and partial least squares considering also their kernelized versions which relation was derived in a similar way as among discriminant analysis methods used in this study. Some preliminary notes are also mentioned in .…”

Section: Methodsmentioning

confidence: 99%

“…Fisher linear discriminant analysis (also the so called Fisher discriminant analysis) and principal component analysis (PCA) supplemented by descriptive statistic outputs. Fisher discriminant analysis was extended by analytical paths of eigenvalues and eigenvectors to originate the other boundary point corresponding to its partial least squares variant derived likewise the connection of canonical correlation analysis and partial least squares (Malec, 2016) with various statistical properties of its boundaries, see e.g. the studies by Wegelin (2000) or Malec (2013).…”

Section: Introductionmentioning

confidence: 99%

Evaluating Competencies of Graduates in Tourism as a Prerequisite for Future Employability

Malec¹,

Kiráľová²

2018

Prague Economic Papers

Self Cite

View full text Add to dashboard Cite

This study attempted to evaluate the differences between bachelor and master requirements in the field of tourism regarding sixteen communication as well as sixteen managerial competencies explored in a wide range of employers in the Czech Republic. The questionnaire survey data were processed using a multivariate approach of the Fisher linear discriminant analysis (LDA) and its partial least squares algorithm (plsLDA) considering two-set cases and the functional relation of both methods depending on a parameter. It was revealed that although the differences between bachelor and master requirements in communication competencies are more pronounced considering the results of the classical approach to discriminant analysis, this is not the case in its partial least squares variant. The differences are to a great extent variable, filling the whole set of individual competencies. However, the most significant competencies demanded from master graduates are: providing a complex view -summary, providing feedback, time management, leadership and planning, considering both boundaries of eigenvalues with corresponding eigenvectors analytical (resp. smooth) path. The practical implication of the results can serve as a target for modern educational systems preparing students for success in the current competitive environment.

show abstract

“…To evaluate economic data, various statistical methods, including multidimensional ones, are commonly employed; see, e.g., Malec (2016) or Malec and Janovský (2019). In this research, standard analytical methods were used to model time series and construct predictions.…”

Section: Introductionmentioning

confidence: 99%

The Czech Republic’s Economy since the 1990s and Its Development Forecasts: Selected Macroeconomic Indicators

Bílková

2023

View full text Add to dashboard Cite

Using a comparative developmental analysis of wage levels and other indicators, the present paper aims to capture the changes in the economy and measurable aspects of living standards in the Czech Republic since the beginning of the then Czechoslovakia’s transformation process launched in the early 1990s. To model the trend, analysis of the relevant time series was performed, exponential smoothing as an adaptive approach to trend modeling being applied. Interpolation and extrapolation criteria were used to verify the suitability of the selected exponential equalization; the Durbin-Watson test, determination index, residual autocorrelation functions, t-tests and Theil inequality coefficient were employed as well. The variables analyzed are the average wage, GDP and the rates of inflation and unemployment, supplemented by the minimum wage and the subsistence minimum. The data used come from the official website of the Czech Statistical Office. They were processed using SPSS and SAS statistical software and a Microsoft Excel spreadsheet. The results show economic downturns and gradual recoveries accompanied by repeated freezing and starting wage growth. High inflation rates in the transition from a centrally planned to a market economy or the GDP decline at the beginning of the financial crisis are evident, the labor market reacting with some delay, as reflected in the rise in the unemployment rate after the currency crisis in the late 1990s, and in average wage trends during the decline of the economic crisis in the early 2010s. The purpose of this research is also to predict further wage level and macroeconomic aggregate developments until 2023. This study aspires to be the starting point for subsequent research, comparing the original forecasts with the reality of further economic development interrupted by the COVID-19 pandemic.

show abstract

Some remarks on the functional relation between canonical correlation analysis and partial least squares

Cited by 9 publications

References 14 publications

Local Food as a Tool of Tourism Development in Regions

Local Food as a Tool of Tourism Development in Regions

Evaluating Competencies of Graduates in Tourism as a Prerequisite for Future Employability

The Czech Republic’s Economy since the 1990s and Its Development Forecasts: Selected Macroeconomic Indicators

Contact Info

Product

Resources

About