Antonio Lucadamo

Principal component multinomial regression and spectrometry to predict soil texture

The mineral particles are classified in different textural classes according to their size. Reflectance spectrometry and spectra can be valid instruments to classify the soils according to their texture. This is possible using different statistical methods, for example, discriminant analysis. However, other multivariate methods, like multinomial logistic regression, can be used, but the presence of multicollinearity among explicative variables could affect the estimation of the parameters. The solution proposed to remedy this problem is an alternative way to apply the multinomial logit model. To evaluate its performances, we compare the results with both classical multinomial logit and discriminant analysis ones. Copyright

A New Scaling Proposal for Handling Ordinal Categorical Variables in Co-inertia (-PLS) Analysis

Abstract In order to investigate the symmetrical relationships between several sets of variables, or regress one or more quantitative response variables on a set of variables of different nature, it is well known that it is necessary to transform non-quantitative variables in such a way that they can be analyzed together with the others measured on an interval scale. This paper suggests a proposal to cope with the problem of the treatment of ordinal qualitative variables in Co-Inertia(-PLS) Analysis. In the literature there are different proposals based on the application of known statistical techniques to quantify ordinal variables. The approach consists in quantifying each non-quantitative variable according to the empirical distributions of the variables involved in the analysis assuming the presence of a continuous underlying variable for each ordinal indicator.

A proposal for handling ordinal categorical variables in co-inertia analysis

This paper is about the problem of the treatment of ordinal qualitative variables in co-inertia analysis. In the literature, there are different proposals based on the application of known statistical techniques to quantify ordinal variables. Here we propose to use a new procedure for the coding considering the empirical distributions of the variables involved in the analysis. We present an application to a real dataset, comparing the results obtained with the different kinds of quantification.

Taxicab Non Symmetrical Correspondence Analysis for the Evaluation of the Passenger Satisfaction

Taxicab Non Symmetrical Correspondence Analysis (TNSCA) is a technique which is more robust than the ordinary Non Symmetrical Correspondence Analysis (NSCA). TNSCA is a variant of the classical Correspondence Analysis (CA) for analyzing the two-way contingency table with a structure of dependence between two variables. In order to overcome the influence due to the presence of the outlier, TNSCA gives uniform weights to all points based on the taxicab singular value decomposition. The visual map constructed by TNSCA offers a clearer perspective than that obtained by correspondence analysis and it may be very useful in evaluating the satisfaction of public transportation passengers.