Using Intelligent Optimisation Methods to Improve the Group Method of Data Handling in Time Series P...
Download
1 / 26

Talk - PowerPoint PPT Presentation


  • 308 Views
  • Updated On :

Using Intelligent Optimisation Methods to Improve the Group Method of Data Handling in Time Series Prediction Maysam Abbod and Karishma Dashpande School of Engineering and Design Brunel University, West London Outline GMDH Genetic Algorithms Particle Swarm Optimisation Financial Data

loader
I am the owner, or an agent authorized to act on behalf of the owner, of the copyrighted work described.
capcha
Download Presentation

PowerPoint Slideshow about 'Talk' - benjamin


An Image/Link below is provided (as is) to download presentation

Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author.While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server.


- - - - - - - - - - - - - - - - - - - - - - - - - - E N D - - - - - - - - - - - - - - - - - - - - - - - - - -
Presentation Transcript
Slide1 l.jpg

Using Intelligent Optimisation Methods to Improve the Group Method of Data Handling in Time Series Prediction

Maysam Abbod and Karishma Dashpande

School of Engineering and Design

Brunel University, West London

Dr M F Abbod


Outline l.jpg
Outline Method of Data Handling in Time Series Prediction

  • GMDH

  • Genetic Algorithms

  • Particle Swarm Optimisation

  • Financial Data

  • Prediction Results

  • Conclusions

Dr M F Abbod


Introduction l.jpg
Introduction Method of Data Handling in Time Series Prediction

  • The GMDH is an algorithm to learn inductively, combinatorial multi-layers for modelling complex systems.

  • The method was introduced by A. G. Ivakhnenko in 1966 and several scholars has since developed the theory GMDH for various applications.

Dr M F Abbod


Slide4 l.jpg
GMDH Method of Data Handling in Time Series Prediction

An important feature of the algorithm GMDH is providing robust polynomial regression models of linear and non-linear systems.

Dr M F Abbod


Principle of selection l.jpg
Principle of Selection Method of Data Handling in Time Series Prediction

Ivakhnenko uses the principles of selectivity - "to get plants, for example, with certain properties, there is the first cross and then the first harvest. Later picks up the best plants and it is the second crossing and the second harvest and thus to find a plant that is desired. "

Dr M F Abbod


Slide6 l.jpg
GMDH Method of Data Handling in Time Series Prediction

GMDH-layers All combinations of inputs are generated and issued the first layer of the network. The outputs of these are classified and then selected for entry into the next layer with all combinations of selected outlets. Only those elements whose performance was acceptable survive to form the next layer. This process is continued as long as each layer (n +1) subsequent produce a better result than the layer (n). When the layer (n +1) is not better as the layer (n), the process is stopped.

Dr M F Abbod


Slide7 l.jpg
GMDH Method of Data Handling in Time Series Prediction

Dr M F Abbod


The choice of plymomial eq l.jpg
The Choice of Plymomial Eq Method of Data Handling in Time Series Prediction

  • GMDHEach layer consists of Polynomial Equation generated from combinations of pairs of inputs. Each node is the way Ivakhnenko polynomial which is a polynomial of the second order: The error we are computed by RMSE and MAPE:

Dr M F Abbod


The coefficients l.jpg
The Coefficients Method of Data Handling in Time Series Prediction

Determining the values that can produce the best adjustment of the equation

Dr M F Abbod


Genetic algorithms l.jpg
Genetic Algorithms Method of Data Handling in Time Series Prediction

It was developed by Goldberg in 1989.

Genetic Algorithms (GAs) are randomised search and optimisation techniques guided by the principles of evolution and natural genetics

Dr M F Abbod


Genetic algorithms11 l.jpg
Genetic Algorithms Method of Data Handling in Time Series Prediction

  • Chromosomes are an encoded representations of the solutions, each gene represents a feature

  • A fitness value that reflects how good it is

  • A crossover mechanism that exchanges portions between strings

  • Mutation plays the role of regenerating lost genetic material

Dr M F Abbod


Particle swarm optimisation l.jpg
Particle Swarm Optimisation Method of Data Handling in Time Series Prediction

Rules of movement – the formulas:

y

x

Dr M F Abbod


The data l.jpg
The Data Method of Data Handling in Time Series Prediction

  • USD2EURO from 29 Sept, 2004 to 5 Oct, 2007.

  • GBP2USD from 29 Sept, 2004 to 5 Oct, 2007.

  • www.oanda.com

Dr M F Abbod


The data14 l.jpg
The Data Method of Data Handling in Time Series Prediction

  • 2 data sets (GBP2USD & USD2EUR)

  • 120 Data points

  • 100 for training

  • 20 for testing

Dr M F Abbod


Training data performance l.jpg
Training Data Performance Method of Data Handling in Time Series Prediction

Dr M F Abbod


Slide16 l.jpg
GMDH Method of Data Handling in Time Series Prediction

GMDH predictions on testing set for

(a) USD2EUR, and (b) GBP2USD

Dr M F Abbod


Pso gmdh gbest l.jpg
PSO-GMDH (gbest) Method of Data Handling in Time Series Prediction

PSO-GMDH gbest model predictions on testing set for

(a) USD2EUR and (b) GBP2USD

Dr M F Abbod


Pso gmdh lbest l.jpg
PSO-GMDH (lbest) Method of Data Handling in Time Series Prediction

PSO-GMDH lbest model predictions on testing set for

(a) USD2EUR and (b) GBP2USD

Dr M F Abbod


Ga gmdh l.jpg
GA-GMDH Method of Data Handling in Time Series Prediction

GA-GMDH predictions on testing set for

(a) USD2EUR, and (b) GBP2USD

Dr M F Abbod


Ga pso gmdh l.jpg
GA-PSO-GMDH Method of Data Handling in Time Series Prediction

GA-PSO-GMDH predictions on testing set for

(a) USD2EUR and (b) GBP2USD

Dr M F Abbod


Testing data performance l.jpg
Testing Data Performance Method of Data Handling in Time Series Prediction

Dr M F Abbod


Usd2eur l.jpg
USD2EUR Method of Data Handling in Time Series Prediction

Dr M F Abbod


Gbp2usd l.jpg
GBP2USD Method of Data Handling in Time Series Prediction

Dr M F Abbod


Performance improvements l.jpg
Performance Improvements Method of Data Handling in Time Series Prediction

Dr M F Abbod


Computational requirements l.jpg
Computational Requirements Method of Data Handling in Time Series Prediction

Dr M F Abbod


Conclusions l.jpg
Conclusions Method of Data Handling in Time Series Prediction

  • Improvements can be achieved

  • Model Complexity and Computational burden

  • Parallel Processing (Matlab: Parallel Computing Toolbox)

  • Other data sets

Dr M F Abbod


ad