Neural Networks
Neural Net Models for Teachers and Students
Artificial neural networks have revolutionized the way researchers solve
many complex and real-world problems in image processing, engineering,
science, economics and finance. The Wolfram Language includes a wide range
of state-of-the-art integrated machine learning capabilities, including a
neural network framework,
and the
Wolfram Neural Net Repository
contains neural network models available for training complicated nets
on real-world data. This Neural Networks add-on package is intended for
teaching and investigating simple neural net models on small datasets.
The Neural Networks package gives teachers and students tools to train,
visualize and validate simple neural network models. It supports a
comprehensive set of neural network structures, including radial
basis function, feedforward, dynamic, Hopfield, perceptron, vector
quantization, unsupervised and Kohonen networks. It implements
training algorithms such as Levenberg–Marquardt, Gauss–Newton
and steepest descent. Neural Networks also includes special
functions to address typical problems in data analysis, such
as function approximation, classification and detection,
clustering, nonlinear time series and nonlinear system
identification problems.
The Neural Networks package features palettes that facilitate the
input parameters for the analysis, evaluation and training of your
data. The documentation contains a number of detailed examples that
demonstrate different neural network models and algorithms. You can
solve many problems simply by applying the example commands to your
own data. The Neural Networks package also provides numerous options
to modify the training algorithms. The default values have been set
to give good results for a large variety of problems, allowing you
to get started quickly using only a few commands. As you gain
experience, you will be able to customize the algorithms to
improve the performance, speed and accuracy of your neural
network models.
The package comes with electronic documentation that contains a
number of detailed examples that demonstrate the use of the different
neural network models, making the Neural Networks package an excellent
teaching tool either for independent study or for use in neural
network courses.
About the Developer
Jonas Sjöberg is a professor and the head of the mechatronic research
group at Chalmers University. Dr. Sjöberg's research involves mechatronic-related
fields, such as signal processing and control. Within these fields, he focuses
on model-based methods, simulations, system identification and optimization
for design and product development of mechatronic systems.
Product Support
The Neural Networks add-on is developed and supported by Dr. Jonas Sjöberg.
Dr. Jonas Sjöberg
email: jonas.sjoberg@chalmers.se
Neural Networks 1.4 requires Mathematica 11 or greater and is available for all
Mathematica platforms.
|
