Neuroph 2.5b

Download now 1.60MB

View screenshots (4)

Your rating

(0) 0 Votes

Our rating

Date added

January 22, 2011

Developer

Zoran Sevarac | (more products)

All time downloads

580

License

Free

Platform

Windows

Description of Neuroph 2.5b

Lightweight Java neural network framework

Neuroph is lightweight Java neural network framework which can be used to simulate common neural network architectures, It has fast learning curve with small number of basic classes which correspond to basic NN concepts.If you are beginner with neural networks, you just want to try how they work without going into complicated theory and implementation, or you need them for your research project the Neuroph is for you.

Neuroph supports the following NN architectures: Adaline, Perceptron, Multi Layer Perceptron, Hopfield, Kohonen, NFR (Neuro Fuzzy Reasoner).

Neuroph is a small, well documented, easy to use, but very flexible neural network framework.

Requirements for Neuroph 2.5b

Java

Editor's review for Neuroph

Traditional computer programs function on linear logic which dates back to John von Neumann who was a Hungarian-born American mathematician regarded as the father of digital computing. Linear logic is essentially a list of instructions based around "IF this IS TRUE THEN do activity A ELSE do activity B", for example: "IF today IS John's birthday THEN wish him Happy Birthday ELSE try again tomorrow" which works well for situations with known inputs and outputs.

Computing tasks involving some degree of prediction or object recognition would be laboriously hard or impossible to program efficiently this way. A new method of computation is required to solve this dilemma. Fortunately, every person carries around a perfect solution: the human brain. The human brain is capable of perceiving an object (courtesy of the eye), identifying straight lines, curves, colours and other identifying characteristics then comparing that data with known objects and determining what type of object is being viewed. These characteristics are stored in what are called neurons which are interconnected by neural pathways to form a neural network.

Artificial neural networks are a simplified, computer based version of this methodology which can be used to solve scenarios that would be almost impossible to resolve using linear logic. Popular applications of neural networks are Optical Character Recognition (OCR) which attempts to convert images containing text into editable documents, stock market prediction, and image recognition applications such as fingerprint and facial recognition.

The Neuroph JAVA Framework simplifies the development of neural networks by providing a well designed, open source Java library with small number of basic classes corresponding to basic neural network concepts. A GUI tool is included that supports creating, training and saving neural networks. If you are a new student of neural networks or you are curious and want to discover how they work without getting bogged down in complicated theory and implementation, then Neuroph is a good choice for you. In a nutshell, artificial neural networks are no longer alien to you with the help of the Neuroph JAVA Framework.

Artificial neural networks is made simple with the Neuroph JAVA Framework.