[ACCEPTED]-Open-source .NET neural network library?-neural-network
Best .NET neural network library is AForge
Encog is a free open source neural network 1 API for both Java and DotNet.
NeuronDotNet It is not as broad in its scope as AForge, but 14 its exclusive focus on NN makes it maybe 13 more readily accessible, and possibly more 12 featured in this area.
Update: (December 2012)
NeuronDotNet's 11 old repository at http://neurondotnet.freehostia.com is defunct.
The most 10 recent source code, Version 3.0 is available on sourceforge, but 9 this project is effectively inactive. Apparently, its original developer, and 8 sole contributor, Vijeth Dinesha, has stopped 7 updating and improving this code base. Thank 6 you, Vijeth! Maybe someone will, some day, pick-up 5 from where you left. Indeed, NeuronDotNet 4 is a relatively mature framework and its 3 simplicity and single focus make it attractive, compared 2 with other frameworks where Neural Nets 1 are just one feature/module therein.
In case this helps anybody else, MS SQL 3 has pretty robust built-in Neural Network support as part of Data Mining Analysis Services.
I know MS SQL isn't open 2 source, but odds are if you are using .NET 1 your data may already be in SQL.
Here is a list of F# related resources for 23 open source .NET machine learning.
Frameworks 22 available on NuGet: (The content below was 21 taken directly from the above referenced 20 URL for persistence purposes.)
Accord.MachineLearning 19 - Contains Support Vector Machines, Decision 18 Trees, Naive Bayesian models, K-means, Gaussian 17 Mixture models and general algorithms such 16 as Ransac, Cross-validation and Grid-Search 15 for machine-learning applications. This 14 package is part of the Accord.NET Framework.
Encog 13 Machine Learning Framework - An advanced 12 neural network and machine learning framework. Encog 11 contains classes to create a wide variety 10 of networks, as well as support classes 9 to normalize and process data for these 8 neural networks. Encog trains using multithreaded 7 resilient propagation. Encog can also make 6 use of a GPU to further speed processing 5 time. A GUI based workbench is also provided 4 to help model and train neural networks.
Numl 3 - A machine learning library intended to 2 ease the use of using standard modeling 1 techniques for both prediction and clustering
From what I can see: FANN fits both of the categories.
Download Weka and convert from the jar file 4 (Java bytecode) to .NET Framework (managed) dll 3 through IKVM. Weka is aknowledged by many as 2 a very good open-source library for machine 1 learning. It includes neural networks.
https://dotnet.github.io/infer/ Infer.NET is a framework for running Bayesian 1 inference in graphical models.
More Related questions