Page 188 - DCAP208_Management Support Systems

Unit 11: Neural Networks

Network Structure Notes

Each ANN is composed of a collection of neurons, grouped in layers. A typical structure is
shown in Figure 11.3. Note the three layers: input, intermediate (called the hidden layer), and
output. A hidden layer is a layer of neurons that takes input from the previous layer and
converts those inputs into outputs for further processing. Several hidden layers can be placed
between the input and output layers, although it is quite common to use only one hidden layer.
In that case, the hidden layer simply converts inputs into a nonlinear combination and passes
the transformed inputs to the output layer. The most common interpretation of the hidden layer
is as a feature extraction mechanism. That is, the hidden layer converts the original inputs in the
problem into some higher level combinations of such inputs.
Figure 11.3: Neural Network with One Hidden Layer

Source: http://www70.homepage.villanova.edu/matthew.liberatore/Mgt2206/turban_online_ch06.pdf
Like a biological network, an ANN can be organized in several different ways (i.e., topologies
or architectures); that is, the neurons can be interconnected in different ways. Therefore, ANN
appear in many configurations called architectures. When information is processed, many of the
processing elements perform their computations at the same time. This parallel processing
resembles the way the brain works, and it differs from the serial processing of conventional
computing.

11.1.4 Network Information Processing

Once the structure of a neural network is determined, information can be processed. We now
present the major concepts related to the processing.
Inputs: Each input corresponds to a single attribute.

Example: If the problem is to decide on approval or disapproval of a loan, some attributes
could be the applicant’s income level, age, and home ownership.

The numeric value, or representation, of an attribute is the input to the network. Several types
of data, such as text, pictures, and voice, can be used as inputs. Preprocessing may be needed to
convert the data to meaningful inputs from symbolic data or to scale the data.

LOVELY PROFESSIONAL UNIVERSITY 181