Transcript WEKA - The A Group of BI's Blog

Artificial Neural Network Building
Using WEKA Software
Arief Rakhman
Goeij Yong Sun
Rama Catur
Outline Cloud
ANN
MLP
WEKA
WEKA
Main Features
MLP in WEKA
Practice!
ANN
(Artificial Neural Network)
a set of connectionist models inspired in the behavior of the human brain
ANN (2)
• Artificial Neural Network is a mathematical model or
computational model that tries to simulate the structure
and/or functional aspects of biological neural networks.
• ANN consists of an interconnected group of artificial neurons
and processes information using a connectionist approach to
computation [10].
• ANN is an adaptive system that can change structures itself
based information that affect the process during computation
of connectioning approach.
• ANN is kind of non-linear statistical data modeling tool. It
usually used with complex model or to find pattern of data.
MLP
(Multilayer Perceptron)
the most popular ANN architecture, where neurons are grouped in layers and
only forward connections exist [1]
weight
perceptron
x1
output
input
xn
hidden layer
MLP (2)
• MLP provides a powerful base-learner, with
advantages such as nonlinear mapping and
noise tolerance,
• Increasingly used in Data Mining due to its
good behavior in terms of predictive
knowledge [2]
WEKA
• A kind of bird in Hamilton, New Zealand
• Waikato Environment for Knowledge Analysis
• Collection of machine learning algorithms and data processing
tools implemented in Java. Released under the GPL
• Have been developed since 1993
• Support for the whole process of experimental data mining :
– Preparation of input data
– Statistical evaluation of learning schemes
– Visualization of input data and the result of learning
• Used for education, research and applications
WEKA Main Features
•
•
•
•
49 data preprocessing tools
76 classification/regression algorithms (including MLP)
8 clustering algorithms
15 attribute/subset evaluators + 10 search algorithms for
feature selection
• 3 algorithms for finding association rules
• 3 graphical user interfaces
– “The Explorer”
(exploratory data analysis)
– “The Experimenter”
(experimental environment)
– “The KnowledgeFlow”
(new process model interface)
MLP in WEKA
• A Classifier function that uses
backpropagation algoritm to classify instances
• The network can also be monitored and
modified during training time
• The nodes in this network are all sigmoid
(except for when the class is numeric in which
case the the output nodes become
unthresholded linear units)
Practice
• Let’s learn by doing!
References
[1] A Abraham. (2004). Meta learning evolutionary artificial neural networks. In
Neurocomputing 56 (p. 1–38).
[2] D.H. Ackley, M.L. Littman. (1994). A case for Lamarckian Evolution. MA: AddisonWesley (p. 3–10)
[3] Rochaa, M., Cortezb, P., & Nevesa, J. (May 22, 2007). Evolution of Neural Networks
for Classification and Regression. Retrieved from sciencedirect.com:
http://www.sciencedirect.com/
science?_ob=MImg&_imagekey=B6V10-4NSWYYK-5-F&_cdi=5660&_user=
8487756&_orig=search&_coverDate=10%2F31%2F2007&_sk=999299983&view=
c&wchp=dGLbVlz-zSkWz&md5=e37e702aff003293e8fdb91aadcbf9b6&ie=
/sdarticle.pdf
References (2)
Eibe Frank. Sourceforge.net. Retrieved November 4, 2009 from
http://prdownloads.sourceforge.net/weka/weka.ppt OR
https://sourceforge.net/projects/weka/files/documentation/Initial%20upload%20
and%20presentations/weka.ppt/download