COMP652_Project[1]

COMP652_Project[1] - COMP 652 Final Project By Jules...

Info iconThis preview shows pages 1–8. Sign up to view the full content.

View Full Document Right Arrow Icon
COMP 652 Final Project By Jules Fakhoury December 15, 2010 12/15/2010 1
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
High fluctuation and irregularity in the data Artificial Neural Network (ANN) Outperforms most statistical methods Based on the empirical risk minimization which minimizes the error on training data Suffers from over fitting and the large number of parameters 12/15/2010 2
Background image of page 2
Support Vector Machine (SVM) Novel supervised learning method Based on the structural risk minimization which minimizes an upper bound of generalization error Less prone to over fitting Different SVM formulation Support Vector Classification Support Vector Regression 12/15/2010 3
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Given {(x 1 ,y 1 ),…,( x m ,y m )}, we want to find a function f(x) that has a most ℇ deviation for the targets y i => -SVR 12/15/2010 4
Background image of page 4
Graphs taken from [1] 12/15/2010 5
Background image of page 5

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
SVM tries to computes: y=f(x)=w. φ (x)+b by minimizing: The approximate function can be written as: 12/15/2010 6
Background image of page 6
Given the historical closing daily prices of the S&P/TSX 0 index, we want to build a model that could predict future values. 12/15/2010
Background image of page 7

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Image of page 8
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 03/10/2011 for the course COMP 652 taught by Professor Preicup during the Fall '07 term at McGill.

Page1 / 15

COMP652_Project[1] - COMP 652 Final Project By Jules...

This preview shows document pages 1 - 8. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online