Assignment Title.

Added on - 16 Sep 2019

  • Dissertation

    type

  • 1

    pages

  • 414

    words

  • 138

    views

  • 0

    downloads

Showing pages 1 to 1 of 1 pages
Assignment TitleName & student no.1INTRODUCTIONConcise description of task asked and yourapproach(es).2BACKGROUND RESEARCHEvolutionary computing has been used in datamining in many ways. Give a brief summary of someexamples you can find, connecting them to yourassignment perhaps too. Use clear referencing andavoid including material directly unless carefullyindicated as from another source. Half to a wholepage expected.3EXPERIMENTATION3.1 Data Set 1Describe your basic algorithm so anyone who knowsabout basic GAs could hopefully repeat what youhave done. Describe the representation, theparameters, the fitness calculation, etc.Present your results as averaged behaviour overmore than one run, eg (graph not of assignment):Figure 1: Initial performance on dataset 1.Show the effects of varying parameters and givepotential explanations as to why thebehaviour/performance changes. Eg vary themutation rate or population size.3.2 Data Set 2If not in the previous section, this should includehow you allow generalisation in the representation.Again, show results from varying parameters/aspectsand explain them. The use of evolution to learn howmany rules to use would be good to include here and/or other modifications.3.3 Data Set 3 (and UCI data)Clearly describe how you changed the representationto deal with real-valued data. And then presentresults from its use, with graphs and explanations.For the very keen, it would be great to seecomparative performance on the UCI dataset tooand/or the use of another representation scheme, eg,evolving neural networks, or use of Weka, etc.4CONCLUSIONSConcise summary of what you found and learned.Identification of ways you might do thingsdifferently next time, and why.REFERENCESBernado Mansilla, E. & Garrell, J. (2003)Accuracy-BasedLearning Classifier Systems: Models, Analysis andApplications to Classification Tasks.EvolutionaryComputation11(3): 209-238.And please include either a link to your sourcecode so that more than one person can access itover the next few months or as a full listing in anAppendix. The second marker and externalexaminer need to be able to see what you actuallyproduced.
desklib-logo
You’re reading a preview
card-image

To View Complete Document

Become a Desklib Library Member.
Subscribe to our plans

Unlock This Document