Synapse
Category Intelligent Software>Neural Network Systems/Tools
Abstract Synapse is one of the most advanced adaptive and neural systems development environment on the market.
It allows you to design, train and deploy adaptive systems, such as neural networks, to any .NET Framework or .NET Compact Framework compatible platform.
Synapse combines an easy-to-use graphical user interface with advanced algorithms for adaptive systems as well as a component-based design approach that allows integration with existing systems.
Popular uses of adaptive systems include Function Modeling, Classification, System Identification, Clustering, Features Extraction, Dynamic Filtering and Anomaly Detection.
Synapse 1.5 Main highlights:
1) Full multicore support for optimizers.
2) A new Fully Forward Connected Perceptron Array (FFCPA) advanced neural network (NN) block.
3) New non-linear gradient optimization algorithms.
4) Improved SQL interoperability and data loading performance improvement.
5) Control system customization (for deployment).
6) Many performance improvements and bug fixes.
Products features/capabilities include:
The Key to Your Data -- Discover a world of possibilities with your data using Synapse. Advanced visualizers allow you to study your data up close. Hierarchical data mining lets you discover patterns and relations hidden in the numbers.
Use advanced filters to non-destructively process, shape and refine your data. Synapse features a wide array of filters ranging from outlier removal and wavelet transforms to equation based expression and selection filtering.
Component Based Design -- In Synapse you are Not limited to just a few conventional models. Model design in Synapse follows a component based data-flow principle. You freely connect different types of components to form a common topology.
Flexible Topologies -- Use pre-defined topology snippets from Synapse’s library for rapid construction of models. You can also create your own snippets for future use. The Extensible Markup Language (XML) based snippets can readily be shared with colleagues and co-workers.
The Avant-Garde -- Take advantage of some of the most cutting-edge technologies for adaptive systems. Synapse implements everything from static neural networks, support vector machines (SVMs), neuro-fuzzy systems and Bayesian classifiers to fully recurrent systems trained with Dynamic XProp (The Dynamic XProp is a bi-directional component based signal control system.
It directs all data to components, propagators and update rules. The dynamic version synchronizes the bi-directional flow in time so that recurrent systems with local memory can be updated correctly in both passes).
The optimization algorithms used for system adaptation range from higher order gradient methods to competitive learning and particle swarm optimization;
[The Particle Swarm Optimization (PSO) batch processor uses a type of swarm intelligence to search for optimal parameters in a system.
Usage - Optimizers are used to search for optimal parameters to components that support optimization. The supervised family of optimizers uses a cost function component (like the Delta Terminator) for a fitness estimate].
Post-Processing -- Enjoy advanced post-adaptation analysis tools. Test, verify and understand the systems you have adapted.
Understand how modifications of system inputs affect the outputs using non-linear sensitivity analysis. Test your model in real-time through a simple interface. Verify the performance of your adapted system with confidence and error metrics.
Deployment -- Export your systems for use in real-world applications. Synapse is built around Microsoft’s .NET technology. Once created, an adaptive system can be natively deployed as a single .NET component. It can then be integrated into other applications.
Go mobile with your systems. Solutions deployed from Synapse are compatible with the .NET Compact Framework, allowing a low footprint deployment to devices such as Pocket PCs and Smart phones.
Expansions -- Tailor Synapse to your needs. In Synapse everything is a plug-in, so you can expand Synapse using standard tools (C#, VB.NET). You can easily create new components or extend existing ones.
Non Linear Workflow -- Work smart in Synapse. Take advantage of the unrestricted non linear workflow. Use the information from post-processing and send it back to pre-processing for analysis.
Track your samples through the system and see how specific input data performs. In Synapse you can always jump to a previous stage and improve your system.
Always Up to Date -- Synapse is updated on a regular basis to bring you the latest technology. The automatic update engine makes sure that your own copy is always up to date.
Peltarion is dedicated to ensuring that Synapse is one of the first software products to use cutting-edge technologies today as well as in the future.
System Requirements
Microsoft Windows Platform
- CPU: Pentium 4 at 1 GHz. For optimal experience and performance the manufacturer recommends at least a dual core CPU.
- RAM: 512 MB (1+ GB recommended)
- Monitor: 1200×1024 (Recommended)
- Disk Space: 25 MB (an additional 280 MB (x86), 610 MB (x64) is needed If .NET Framework is Not already installed.)
- OS: Windows 2000, XP, 2003 Vista or 7.
Manufacturer
- Peltarion HB
- Sveavägen 52
- 111 34 Stockholm
- Sweden
- info@peltarion.com
- sales@peltarion.com
- system-solutions@peltarion.com
Manufacturer Web Site Synapse
Price 999 Euros
G6G Abstract Number 20113R
G6G Manufacturer Number 102120