PSO in ANN, SVM and data clustering

PSO in ANN, SVM and data clustering

For access to this article, please select a purchase option:

Buy chapter PDF
(plus tax if applicable)
Buy Knowledge Pack
10 chapters for £75.00
(plus taxes if applicable)

IET members benefit from discounts to all IET publications and free access to E&T Magazine. If you are an IET member, log in to your account and the discounts will automatically be applied.

Learn more about IET membership 

Recommend Title Publication to library

You must fill out fields marked with: *

Librarian details
Your details
Why are you recommending this title?
Select reason:
Swarm Intelligence - Volume 1: Principles, current algorithms and methods — Recommend this title to your library

Thank you

Your recommendation has been sent to your librarian.

In this chapter, one gives an introduction to different kinds of particle swarm optimization (PSO) algorithms. One also introduces artificial neural networks (ANNs), support vector machines (SVMs) and evolutionary computing to show how PSO may be used to determine optimal parameters using an ANN or SVM regime, for classification of DNA strings. In addition, PSO is used in the design of an SVM-based clustering algorithm. Ant colony optimization (ACO) algorithms are also introduced in the chapter. Using ACO algorithms has been of great success to solve many discrete optimization and non-deterministic polynomial (NP)-hard problems, for instance the travelling salesman problem. The behaviour of ants is also used, for instance, to design an algorithm for data clustering. We want to develop later a similar application based on this clustering algorithm and compare it with the SVM one using PSO.

Chapter Contents:

  • Abstract
  • 18.1 Introduction
  • 18.2 Global particle optimization
  • 18.3 Local particle optimization
  • 18.4 PSO variations
  • 18.4.1 Clamping
  • 18.4.2 Inertia weight
  • 18.4.3 Constriction coefficient
  • 18.4.4 Algorithmic aspects
  • Velocity models
  • Initialization
  • Termination
  • 18.5 PSO parameters
  • 18.6 Evolutionary computing and PSO
  • 18.6.1 Mutation
  • 18.6.2 Crossover
  • 18.6.3 Summary
  • 18.6.4 A selection-based PSO algorithm
  • 18.6.5 Replacement
  • 18.7 A multi-phase PSO
  • 18.8 Custom PSO
  • 18.9 Artificial neural network
  • 18.9.1 The biological neuron
  • 18.9.2 The multi-layer perceptron
  • The artificial neuron
  • The layers
  • The backpropagation algorithm
  • 18.10 Support vector machine
  • 18.10.1 The Kernel
  • 18.10.2 Definition of Kernels
  • 18.10.3 Quadratic programming
  • 18.10.4 KKT conditions
  • 18.11 DNA classification
  • 18.11.1 Presentation of the problem
  • 18.11.2 DNA recognition theory
  • 18.11.3 Feature extraction
  • Normalization
  • 18.11.4 Use of PSO
  • 18.11.5 Experiments
  • Results
  • 18.12 Support vector clustering
  • 18.12.1 Theory
  • 18.12.2 SVC–PSO relation
  • 18.12.3 Data set
  • 18.12.4 The experiment
  • 18.13 Ant colony systems
  • 18.13.1 The pheromone
  • 18.13.2 Optimization: travelling salesman problem
  • 18.14 Ant colonies clustering
  • References

Inspec keywords: neural nets; evolutionary computation; particle swarm optimisation; travelling salesman problems; support vector machines; computational complexity; pattern clustering

Other keywords: SVM regime; data clustering; DNA strings; optimal parameters; artificial neural networks; particle swarm optimization algorithms; evolutionary computing; PSO; ACO algorithms; clustering algorithm; ANN; ant colony optimization algorithms; discrete optimization; support vector machines; nondeterministic polynomial-hard problems; travelling salesman problem

Subjects: Neural computing techniques; Optimisation techniques; Knowledge engineering techniques; Combinatorial mathematics; Data handling techniques; Computational complexity

Preview this chapter:
Zoom in

PSO in ANN, SVM and data clustering, Page 1 of 2

| /docserver/preview/fulltext/books/ce/pbce119f/PBCE119F_ch18-1.gif /docserver/preview/fulltext/books/ce/pbce119f/PBCE119F_ch18-2.gif

Related content

This is a required field
Please enter a valid email address