Data Mining and Neural Networks Danny Leung

Embed Size (px)

DESCRIPTION

HHHHH

Citation preview

  • Data Mining and Neural Networks

    Danny LeungCS157B, Spring 2006Professor Sin-Min Lee

  • Artificial Intelligence for Data MiningNeural networks are useful for data mining and decision-support applications.

    People are good at generalizing from experience.

    Computers excel at following explicit instructions over and over.

    Neural networks bridge this gap by modeling, on a computer, the neural behavior of human brains.

  • Neural Network Characteristics

    Neural networks are useful for pattern recognition or data classification, through a learning process.

    Neural networks simulate biological systems, where learning involves adjustments to the synaptic connections between neurons

  • Anatomy of a Neural Network

    Neural Networks map a set of input-nodes to a set of output-nodes

    Number of inputs/outputs is variable

    The Network itself is composed of an arbitrary number of nodes with an arbitrary topology

    Neural Network

    Input 0

    Input 1

    Input n

    ...

    Output 0

    Output 1

    Output m

    ...

  • Biological BackgroundA neuron: many-inputs / one-output unit

    Output can be excited or not excited

    Incoming signals from other neurons determine if the neuron shall excite ("fire")

    Output subject to attenuation in the synapses, which are junction parts of the neuron

  • Basics of a NodeA node is an element which performs a function

    y = fH((wixi) + Wb)ConnectionNode

    fH(x)

    Input 0

    Input 1

    Input n

    ...

    W0

    W1

    Wn

    +

    Output

    +

    ...

    Wb

  • A Simple PreceptronBinary logic application

    fH(x) [linear threshold]

    Wi = random(-1,1)

    Y = u(W0X0 + W1X1 + Wb)

    fH(x)

    Input 0

    Input 1

    W0

    W1

    +

    Output

    Wb

  • Preceptron TrainingIts a single-unit network

    Adjust weights based on a how well the current weights match an objective

    Perceptron Learning Rule

    Wi = * (D-Y).Ii

    = Learning RateD = Desired Output

  • Neural Network LearningFrom experience: examples / training data

    Strength of connection between the neurons is stored as a weight-value for the specific connection

    Learning the solution to a problem = changing the connection weights

  • Neural Network LearningContinuous Learning Process

    Evaluate output

    Adapt weights

    Take new inputs

    Learning causes stable state of the weights

  • Learning PerformanceSupervisedNeed to be trained ahead of time with lots of data

    Unsupervised networks adapt to the inputApplications in Clustering and reducing dimensionalityLearning may be very slowNo help from the outsideNo training data, no information available on the desired outputLearning by doingUsed to pick out structure in the input:ClusteringCompression

  • Topologies Back-Propogated NetworksInputs are put through a Hidden Layer before the output layer

    All nodes connected between layers

    Input 0

    Input 1

    Input n

    ...

    Output 0

    Output 1

    Output o

    ...

    O0

    O1

    Oo

    H0

    H1

    Hm

    ...

    ...

    Hidden Layer

  • BP Network Supervised Training

    Desired output of the training examples

    Error = difference between actual & desired output

    Change weight relative to error size

    Calculate output layer error , then propagate back to previous layer

    Hidden weights updated

    Improved performance

  • Neural Network Topology Characteristics

    Set of inputs

    Set of hidden nodes

    Set of outputs

    Increasing nodes makes network more difficult to train

  • Applications of Neural NetworksPrediction weather, stocks, disease

    Classification financial risk assessment, image processing

    Data association Text Recognition (OCR)

    Data conceptualization Customer purchasing habits

    Filtering Normalizing telephone signals (static)

  • OverviewAdvantagesAdapt to unknown situationsRobustness: fault tolerance due to network redundancyAutonomous learning and generalization

    DisadvantagesNot exactLarge complexity of the network structure

  • Referenced WorkIntro to Neural Networks - Computer Vision Applications and Training Techniques. Doug Gray. www.soe.ucsc.edu/~taoswap/ GroupMeeting/NN_Doug_2004_12_1.ppt

    Introduction to Artificial Neural Networks. Nicolas Galoppo von Borries. www.cs.unc.edu/~nico/courses/ comp290-58/nn-presentation/ann-intro.ppt