Neural Networks are programs designed to simulate the way a simple biological nervous
system is believed to operate. They are based on simulated nerve cells or neurons which
are joined together in a variety of ways to form networks. These networks have the
capacity to learn, memorize and create relationships amongst data. There are many
different types of Neural Networks, each of which has different strengths particular to
their applications. The abilities of different networks can be related to their structure,
dynamics and learning methods.
What can you do with a Neural Network?
In principle, Neural Networks can compute any computable function, thus do everything a
normal digital computer can do.
In practice, Neural Networks are especially useful for classification and function
approximation/mapping problems which are tolerant of some imprecision, which have lots of
training data available, but to which hard and fast rules (such as those that might be
used in an expert system) cannot easily be applied. Almost any mapping between vector
spaces can be approximated to arbitrary precision by feedforward Neural Networks (which
are the type most often used in practical applications) if you have enough data and enough
computing resources.
Neural Networks are, at least today, difficult to apply successfully to problems that
concern manipulation of symbols and memory. And there are no methods for training Neural
Networks that can magically create information that is not contained in the training data.
Categories
NN applications are almost limitless but fall into a few simple categories:
Classification:
Among many applications of the feed-forward ANNs, the classification or prediction
scenario is perhaps the most interesting for data mining. In this mode, the network is
trained to classify certain patterns into certain groups, and then is used to classify
novel patterns which were never presented to the net before. Medical diagnosis, signature
verification, voice recognition, image recognition, property valuation, is only a few
examples of this category.
Interactive demonstrations
Being familiar with Neural Networks will allow you to understand these
demonstrations. Please read the section on About Neural Networks
before running these demonstrations.
[These Java demonstrations brought to you by the Neural
Transmitters behind The Mind and Machine Module.]
Forecasting:
Neural Networks is and can be used to predict all sorts of things. Applications
include future sales,
production requirements, market performance, economic indicators,
energy requirements, medical outcomes,
chemical reaction
products, and weather.
Modeling:
Artificial neural networks are flexible multivariate models that can be applied to
predictive modeling and pattern recognition problems. Process control, systems control,
chemical structures, dynamic systems, signal compression, plastics molding, welding
control and robot control, to name a few, would fall under this category.
Conclusion
Past and Present
The development of true Neural Networks is a fairly recent event, which has been
met with success. Two of the different systems (among the many) that have been developed
are: the basic feedforward Network and the Hopfield Net.
The Future
The future of Neural Networks is wide open, and may lead to many answers and/or
questions. Is it possible to create a conscious machine? What rights do these computers
have? How does the human mind work? What does it mean to be human?