CONICET | Buscador de Institutos y Recursos Humanos

Principal Component Analysis has been widely used in different scientific areas and for different applications and purposes. The versatility of this unsupervised multivariate method, allowed the scientific community to explore its applications and potentialities in different areas of knowledge, where the only limitations are researcher imagination and curiosity. Even when the principle of PCA is always the same in what algorithms and fundamentals concerns, the strategies employed to interpret and elucidate both experimental and theoretical information depend principally on the expertise and needs of each researcher. In this chapter, we will focus on a detailed description of the strategies and methodologies currently used to analyze experimental and theoretical data obtained in different applications. Our main goal is to present a general overview of the versatility, capabilities and strengths of PCA method to elucidate and extract specific information or data structure. Additionally, this chapter also pretends to be a reference for those researchers that are not specialists in the field, but that may use these methodologies to take the maximum advantage from their experimental results. In a second part, we will present different examples of the application of PCA on real experimental data. We will report PCA applications on microbiology, clinical diagnosis and quality control in alcoholic beverages. All the presented cases represent good examples of multivariable data sets due to their complexity and the high amount of information contained in them, providing a good platform to demonstrate the potentiality of the PCA method. In this sense, we will focus on the description of the criteria and methodology used for outlier detection, scores and loadings interpretation, and data pre-treatment.

enviar mensaje