Financial prediction using non linear classification techniques

In this thesis, we explore the ability of statistical classification methods to predict financial events in the bond and stock markets. Our classification methods include conventional Linear Dicriminant Analysis (LDA), and a number of less familiar non-linear techniques such as Probabilistic Neural Network (PNN), Learning Vector Quanization (LVQ), Oblique Classifer (OCI), and Ripper Rule Induction (RRI).

Name dc:type.qualificationname

phd

Level dc:type.qualificationlevel

doctoral

Grantor dc:publisher.institution

City University London

Year dc:date.issued

2001

Author dc:creator

• Albanis, G.T.

• HG Finance