|
Title:
|
BEHAVIOURAL FINANCE AS A MULTI-INSTANCE LEARNING PROBLEM |
|
Author(s):
|
Piotr Juszczak |
|
ISBN:
|
978-972-8924-88-1 |
|
Editors:
|
Ajith P. Abraham |
|
Year:
|
2009 |
|
Edition:
|
Single |
|
Keywords:
|
behavioural finance, multi-instance learning |
|
Type:
|
Full Paper |
|
First Page:
|
27 |
|
Last Page:
|
34 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
In various application domains, including image recognition, text recognition or the subject of this paper, behavioural
finance, it is natural to represent each example as a set of vectors. However, most traditional data
analysis methods are based on relations between individual vectors only. To cope with sets of vectors, instead
of single vector descriptions, existing methods have to be modified. The main challenge is to derivemeaningful
similarities or dissimilarities measures between sets of vectors. In this paper, we derive several dissimilarities
measures between sets of vectors. The derived dissimilarities are used as rudiments of data analysis methods,
such as kernel-based clustering and SVMclassification. The performance of the proposedmethods is examined
on consumer credit cards behaviour problems. These problems are shown to be an example of a multi-instance
learning problems. |
|
|
|
|
|
|
