|
Título:
|
Similarity analysis to aid decision making on NBA Draft
|
|
Autores:
|
Houghton López, Miguel Alejandro
|
|
Tipo de documento:
|
texto impreso
|
|
Fecha de publicación:
|
2020
|
|
Dimensiones:
|
application/pdf
|
|
Nota general:
|
cc_by_nc
info:eu-repo/semantics/openAccess
|
|
Idiomas:
|
|
|
Palabras clave:
|
Estado = No publicado
,
Materia = Ciencias: Informática
,
Tipo = Trabajo Fin de Grado
|
|
Resumen:
|
This work is based on the different statistical studies published by Mock Draft Websites and on webs that store the official statistics of the NBA players. The data associated with NBA players and teams are currently very precious since their correct exploitation can materialize in great economic benefits. The objective of this work is to show how data mining can be useful to help the scouts in this real problem. Scouts participating in the Draft could use the information provided by the models to make a better decision that complements their personal experience. This would save time and money since by simply analyzing the results of the models, teams would not have to travel around the world to find players who could be discarded for the choice. In this work, unsupervised grouping techniques are studied to analyze the similarity between players. Databases with statistics of both current and past players are used. Besides, three different clustering techniques are implemented that allow the results to be compared, adding value to the information and facilitating decision- making. The most relevant result is shown at the moment in which the shooting in the NCAA is analyzed using grouping techniques.
|
|
En línea:
|
https://eprints.ucm.es/id/eprint/62872/1/HOUGHTON_LOPEZ_Similarity_Analysis_to_aid_decision_making_on_NBA_Draft_4398578_723276205.pdf
|
