Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different...

43
Chapter 8 NEURAL NETWORKS FOR DATA MINING

Transcript of Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different...

Page 1: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Chapter 8NEURAL NETWORKS FOR DATA MINING

Page 2: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Learning Objectives• Understand the concept and different types of

artificial neural networks (ANN)• Learn the advantages and limitations of ANN• Understand how backpropagation neural

networks learn • Understand the complete process of using

neural networks• Appreciate the wide variety of applications of

neural networks

Page 3: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Basic Concepts of Neural Networks

• Neural networks (NN) or artificial neural network (ANN)Computer technology that attempts to build computers that will operate like a human brain. The machines possess simultaneous memory storage and works with ambiguous information

• A brain metaphor for information processing. These models are biologically inspired rather than an exact replica of how the brain actually functions.

• Very promising systems:– Ability to learn from the data– Nonparametric nature(no rigid assumptions)– Ability to generalize

Page 4: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Basic Concepts of Neural Networks

• Neural computing An experimental computer design aimed at building intelligent computers that operate in a manner modeled on the functioning of the human brain. See artificial neural networks (ANN)

it is a pattern recognition methodology for machine learning. The resulting model from the neural computing is often called ANN.

• Development history (ANN journey)– McCulloch and Pitts, 1943– 1950s-1960s popular– 1970s-1980s diminished– 1990s-, rational developing.– 2000s-, data mining,

Page 5: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Basic Concepts of Neural Networks • Biological and artificial neural networks – Perceptron (感知器) Early neural network structure that uses no hidden layer – Neurons (神经元)

Cells (processing elements) of a biological or artificial neural network – Nucleus (神经核)

The central processing portion of a neuron – Dendrite (树突)

The part of a biological neuron that provides inputs to the cell– Axon (轴突)

An outgoing connection (i.e., terminal) from a biological neuron– Synapse (突触)

The connection (where the weights are) between processing elements in a neural network

Page 6: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Basic Concepts of Neural Networks

Page 7: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Basic Concepts of Neural Networks

Page 8: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Basic Concepts of Neural Networks

Biological Artificial

Soma Node

Dendrite Input

Axon Output

Synapse Weight

Slow Speed Fast Speed

Many neurons (10e9) Few neurons (10-100s)

Page 9: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Basic Concepts of Neural Networks • Elements of ANN – Topologies (architectures)

The type neurons are organized in a neural network – Backpropagation (反向传播,多馈前向)

The best-known learning algorithm in neural computing. Learning is done by comparing computed outputs to desired outputs of historical cases

1986 年 Rumelhart 和 McCelland小组提出。

Page 10: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Basic Concepts of Neural Networks – Processing elements (PEs)

The PE in a neural network are artificial neurons. each of the neurons receives input, processes them, and deliveries

a single output. Show in figure 8.2.– Network structure (three layers)

1. Input2. Intermediate (hidden layer)3. Output • A hidden layer is a layer of neurons that takes input from the

previous layer and converts those input into output for further processing.

• Several hidden layers can be places between input and output layers.

• Most of hidden layers process mechanisms are feature extraction.

Page 11: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Basic Concepts of Neural Networks

Page 12: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Basic Concepts of Neural Networks – Parallel processing

An advanced computer processing technique that allows a computer to perform multiple processes at once—in parallel

Resembles the way the brain works, it differs from the serial processing of conventional computing.

Page 13: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Basic Concepts of Neural Networks

– Network information processing • Inputs : each input corresponds t a single attribute.

E.g., problem is approval or disapproval for a loan. Numeric value, or representation, of an attribute is the

input to the network. • Outputs : contain the solution of a problem. For

example, loan can be yes or no. Also, numeric value to the output, such as 0, or 1 for yes or no.

The purpose of ANN is to compute the values of the output. Thus, post-processing of output is required .

Page 14: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Basic Concepts of Neural Networks – Network information processing • Connection weights

The weight associated with each link in a neural network model. They are assessed by neural networks learning algorithms

• Summation function or transformation (transfer) function In a neural network, the function that sums and transforms inputs before a neuron fires. The relationship between the internal activation level and the output of a neuron

or j11 iinijii

ni WXYWXY

Page 15: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Basic Concepts of Neural Networks

Page 16: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Basic Concepts of Neural Networks

– Sigmoid (logical activation) function An S-shaped transfer function in the range of zero to one

)1/(1 YT eY

Page 17: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Basic Concepts of Neural Networks – Threshold value

A hurdle value for the output of a neuron to trigger the next level of neurons. If an output value is smaller than the threshold value, it will not be passed to the next level of neurons

– Hidden layer• The middle layer of an artificial neural network that has

three or more layers• Theoretically, there can be 10-100 layers. However, more

than three layers are seldom in commercial software.• The more of hidden layers, the more the time will be used

for training.

Page 18: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Basic Concepts of Neural Networks

• Neural network architectures – Common neural network models and algorithms

include:• Backpropagation • Feedforward (or associative memory)• Recurrent network

Page 19: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Basic Concepts of Neural Networks

Page 20: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Basic Concepts of Neural Networks

Page 21: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Learning in ANN– Learning algorithm ( also, training algorithm)

• The training procedure used by an artificial neural network . Learning algorithm specify the process by which a neural network learns the underlying relationship between input and outputs, or just among the inputs.

• There are hundreds of them. • It can be classified into to categories: Supervised learning and

unsupervised learning.

Page 22: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Learning in ANN

Page 23: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Learning in ANN – Supervised learning

A method of training artificial neural networks in which sample cases (training sets) are shown (teaching ) to the network as input and the weights are adjusted to minimize the error in its outputs. The training set and desired output is iteratively presented to the neural networks. Output of the network in its present form is calculated and compared to the desired output.

Backpropagation learning algorithm: is popular supervised learning. It is an iterative gradient-descent technique designed to minimize an

error function between the actual output of the network and its desired output, as specified in the training data set.

– Unsupervised learningA method of training artificial neural networks in which only input stimuli are shown to the network, it organizes itself internally so that each hidden processing element responds strategically to a different set of input stimuli. No knowledge is supplied about which classification are correct.

So it called self-organizing or clustering its neurons related to the specific desired task.

Page 24: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Learning in ANN Self-organizing

A neural network architecture that uses unsupervised learning Adaptive resonance theory (ART)

An unsupervised learning method created by Stephen Grossberg. It is a neural network architecture that is aimed at being more brain-like in unsupervised mode

Kohonen self-organizing feature maps A type of neural network model for machine learning

Page 25: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Learning in ANN • The general ANN learning process – The process of learning involves three tasks:

1. Compute temporary outputs2. Compare outputs with desired targets3. Adjust the weights and repeat the process

Page 26: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Learning in ANN

Page 27: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Learning in ANN • The general ANN learning process – The process of learning involves three tasks:

1. Compute temporary outputs2. Compare outputs with desired targets3. Adjust the weights and repeat the process

Page 28: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Learning in ANN – Pattern recognition

The technique of matching an external pattern to one stored in a computer’s memory; used in inference engines, image processing, neural computing, and speech recognition (in other words, the process of classifying data into predetermined categories).

Page 29: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Learning in ANN • How a network learns – Learning rate

A parameter for learning in neural networks. It determines the portion of the existing discrepancy that must be offset

– Momentum A learning parameter in feedforward-backpropagation neural networks

Page 30: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Learning in ANN • How a network learns – Backpropagation

The best-known learning algorithm in neural computing. Learning is done by comparing computed outputs to desired outputs of historical cases

Page 31: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Learning in ANN • How a network learns – Procedure for a learning algorithm

1. Initialize weights with random values and set other parameters

2. Read in the input vector and the desired output3. Compute the actual output via the calculations, working

forward through the layers4. Compute the error5. Change the weights by working backward from the

output layer through the hidden layers

Page 32: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Developing Neural Network–Based Systems

Page 33: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Developing Neural Network–Based Systems

• Data collection and preparation – The data used for training and testing must include all the

attributes that are useful for solving the problem • Selection of network structure – Selection of a topology– Topology

The way in which neurons are organized in a neural network

Page 34: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Developing Neural Network–Based Systems

• Data collection and preparation – The data used for training and testing must include all the

attributes that are useful for solving the problem • Selection of network structure – Selection of a topology– Determination of:

1. Input nodes2. Output nodes3. Number of hidden layers4. Number of hidden nodes

Page 35: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Developing Neural Network–Based Systems

Page 36: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Developing Neural Network–Based Systems

• Learning algorithm selection – Identify a set of connection weights that best cover the

training data and have the best predictive accuracy • Network training – An iterative process that starts from a random set of

weights and gradually enhances the fitness of the network model and the known data set

– The iteration continues until the error sum is converged to below a preset acceptable level

Page 37: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Developing Neural Network–Based Systems

• Testing – Black-box testing

Comparing test results to actual results – The test plan should include routine cases as well as

potentially problematic situations– If the testing reveals large deviations, the training set must

be reexamined, and the training process may have to be repeated

Page 38: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Developing Neural Network–Based Systems

• Implementation of an ANN – Implementation often requires interfaces with other

computer-based information systems and user training– Ongoing monitoring and feedback to the developers are

recommended for system improvements and long-term success

– It is important to gain the confidence of users and management early in the deployment to ensure that the system is accepted and used properly

Page 39: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Developing Neural Network–Based Systems

Page 40: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

A Sample Neural Network Project

Page 41: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Other Neural Network Paradigms • Hopfield networks – A single large layer of neurons with total interconnectivity—

each neuron is connected to every other neuron– The output of each neuron may depend on its previous

values– One use of Hopfield networks: Solving constrained

optimization problems, such as the classic traveling salesman problem (TSP)

Page 42: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Other Neural Network Paradigms • Self-organizing networks – Kohonen’s self-organizing network learn in an unsupervised

mode – Kohonen’s algorithm forms “feature maps,” where

neighborhoods of neurons are constructed– These neighborhoods are organized such that topologically

close neurons are sensitive to similar inputs into the model– Self-organizing maps, or self organizing feature maps, can

sometimes be used to develop some early insight into the data

Page 43: Chapter 8 NEURAL NETWORKS FOR DATA MINING. Learning Objectives Understand the concept and different types of artificial neural networks (ANN) Learn the.

Applications of ANN • ANN are suitable for problems whose inputs are both

categorical and numeric, and where the relationships between inputs and outputs are not linear or the input data are not normally distributed