Use this link to cite:
http://hdl.handle.net/2183/26207 Menos é máis: explorando o impacto da selección de características
Loading...
Identifiers
Publication date
Authors
Pichel Bolón, Esteban
Other responsabilities
Enxeñaría informática, Grao en
Journal Title
Bibliographic citation
Type of academic work
Academic degree
Abstract
[Resumo]
Neste proxecto estúdanse diferentes técnicas de selección de características coa intención de
determinar cales dos múltiples métodos existentes na literatura son máis axeitados para un
tipo de problema en concreto, e determinar se algún deles é descartable por obter peores
resultados que realizar unha selección aleatoria á hora de reducir a dimensionalidade dos
problemas. Para a súa realización farase uso dun extenso número de conxuntos de datos
que nos permitan traballar sobre unha gran variedade dos problemas existentes no mundo
real cos que nos atopamos neste ámbito, reducindo a súa dimensionalidade e levando a cabo
a clasificación correspondente para obter os resultados que, xunto cos test estadísticos, nos
permitirán sacar conclusións sobre as cuestións plantexadas.
[Abstract] In this project we study different feature selection techniques with the aim of determining which of the multiple methods in the literature are the best suited for a particular type of problem, and determining whether any of them are disposable because of obtaining worse results than a random selection to reduce the dimensionality of the problems. To accomplish this objective we will use an extensive number of data sets that allow us to work on a wide variety of problems from the real world that need to be dealt with in this field. We will reduce the dimensionality of the data sets and carry out the corresponding classification process to obtain the results that, along with statistical tests, will allow us to draw conclusions about the issues raised.
[Abstract] In this project we study different feature selection techniques with the aim of determining which of the multiple methods in the literature are the best suited for a particular type of problem, and determining whether any of them are disposable because of obtaining worse results than a random selection to reduce the dimensionality of the problems. To accomplish this objective we will use an extensive number of data sets that allow us to work on a wide variety of problems from the real world that need to be dealt with in this field. We will reduce the dimensionality of the data sets and carry out the corresponding classification process to obtain the results that, along with statistical tests, will allow us to draw conclusions about the issues raised.







