Document type
Bachelor thesisPublication date
Publication license
Please use this identifier to cite or link to this item: https://hdl.handle.net/2445/171924
Estudi de classificació a partir de tècniques discriminants
Journal Title
Authors
Director/Tutor
Journal ISSN
Volume Title
Related resource
Abstract
[cat] La necessitat d’identificar característiques que ens permetin diferenciar dos o més grups cada
cop és més freqüent. L’anàlisi discriminant consisteix a analitzar aquestes característiques que
ajuden a la classificació de dos o més grups. L’objectiu d’aquest estudi és arribar a una bona
classificació d’una base de dades que té com a observacions les reserves d’un complex hoteler
diferenciat en tres seccions diferents. Aquestes tres seccions de l’hotel seran els nostres grups
de la variable dependent. Els mètodes de classificació amb els quals volem aconseguir una
assignació correcta de reserves noves són l’anàlisi discriminant lineal, l’anàlisi discriminant
quadràtic i l’anàlisi k-nn veïns més pròxims. Seguidament validarem aquests resultats
estudiant la capacitat predictiva. D’aquesta forma arribarem a veure amb quin mètode
obtenim la taxa d’encerts de classificació més elevada.
[eng] The necessity of identifying characteristics that allow us to distinguish two or more groups of individuals is becoming more common. The discriminant analysis consists in analyze these characteristics that you can use at the time of classifying in two or more groups. The goal of this study is to get a good classification of a data base that contain observations of reservations of a hotel complex differentiated in three different sections. These sections will be the groups of our dependent variable. The classification methods with which we want to achieve the classification are linear discriminant analysis, quadratic discriminant analysis and k-nn nearest neighboring analysis. Then, we will validate these results by studying the predictive ability of those methods. In this way, we will be able to see with which method we obtain the highest success rate of classification.
[eng] The necessity of identifying characteristics that allow us to distinguish two or more groups of individuals is becoming more common. The discriminant analysis consists in analyze these characteristics that you can use at the time of classifying in two or more groups. The goal of this study is to get a good classification of a data base that contain observations of reservations of a hotel complex differentiated in three different sections. These sections will be the groups of our dependent variable. The classification methods with which we want to achieve the classification are linear discriminant analysis, quadratic discriminant analysis and k-nn nearest neighboring analysis. Then, we will validate these results by studying the predictive ability of those methods. In this way, we will be able to see with which method we obtain the highest success rate of classification.
Description
Treballs Finals de Grau en Estadística UB-UPC, Facultat d'Economia i Empresa (UB) i Facultat de Matemàtiques i Estadística (UPC), Curs: 2019-2020. Tutor: Mireia Besalú Mayol
Subject (English)
Citation
Citation
MENA GUIX, Silvia. Estudi de classificació a partir de tècniques discriminants. [consulted: 10 of August of 2026]. Available at: https://hdl.handle.net/2445/171924