شبکه ی باور عمیق یکی از مدل های پر کاربرد در یادگیری ژرف هست،
 بسته به فرض های متفاوت در مورد نحوه ی اتصال این یاخته های عصبی ، مدل ها و ساختار های مختلفی در این حوزه پیشنهاد و بررسی شده‌اند ، هرچند که این مدل ها به صورت طبیعی در مغز انسان وجود ندارد و مغز انسان پیچیدگی های بیشتری را دارا است. این مدل ها نظیر شبکه عصبی عمیق (به انگلیسی: deep neural network) ، شبکه عصبی پیچیده (به انگلیسی: convolutional neural network) ، شبکه باور عمیق (به انگلیسی: deep belief network) پیشرفت های خوبی را در حوزه های پردازش زبان‌های طبیعی ، پردازش تصویر ایجاد کرده اند…

کی از بهترین کتاب های  که امروزه خیلی سر صدا کرده تو دنیا و هنوز نسخه اصلیش نیومده و صرفا در حد Draft هستش، کتاب آقای bengio  هستش که انتشارات دانشگاه  MIT داره اونو Release میکنه. پیشنهاد بنده این هست که این کتاب رو از اول بخونین. باید بگم این کتاب صرفا نسخه آنلاینش هست و هنوز براش نسخه PDF بیرون ندادن.

http://www.deeplearningbook.org/

یک مقاله مروری خوب deep learning از مجله nature

شما می توانید مقاله مقاسیه بهترین ابزارهای یادگیری عمیق را از این جا مقالعه کنید

https://github.com/zer0n/deepframeworks#cross-platform

Theano در حقیقت یک کتابخانه برای  python هست
و از طرفی هم Caffe با C++ هست ولی برای Python هم رابط داره و به نظر یکم فراگیرتره و مدلهای آماده بیشتری داره. البته برای بازکردن مدل های Caffe توی کتابخونه های دیگه، توابع آماده هست که می تونید استفاده کنید

چارچوب ( Framework ) نرم افزاری یادگیری ماشین و یادگیری ژرف برای C# و .NET

http://accord-framework.net/

کتابخانه یادگیری ژرف برای جاوا
http://deeplearning4j.org/

ConvNetJS is a Javascript library for training Deep Learning models (Neural Networks) entirely in your browser. Open a tab and you’re training. No software requirements, no compilers, no installations, no GPUs, no sweat.
http://cs.stanford.edu/people/karpathy/convnetjs/

لیستی از 15 کتابخانه یادگیری عمیق

1. Theano is a python library for defining and evaluating mathematical expressions with numerical arrays. Theano supports related frameworks such as Keras, Pylearn2, Lasagne & Blocks.

2. Caffe is a deep learning framework made with expression, speed, and modularity in mind. It is developed by the Berkeley Vision and Learning Center (BVLC) and by community contributors. Google’s DeepDream is based on Caffe Framework. This framework is a BSD-licensed C++ library with Python Interface.

3. nolearn contains a number of wrappers and abstractions around existing neural network libraries, most notably Lasagne, along with a few machine learning utility modules.

4. Chainer bridge the gap between algorithms and implementations of deep learning. Its powerful, flexible and intuitive and is considered as the flexible framework for Deep Learning.

5. deepnet is a GPU-based python implementation of deep learning algorithms like Feedforward Neural Nets, Restricted Boltzmann Machines, Deep Belief Nets, Autoencoders, Deep Boltzmann Machines and Convolutional Neural Nets.

6. ConvNet Convolutional neural net is a type of deep learning classification algorithms, that can learn useful features from raw data by themselves and is performed by tuning its weighs.

7. SINGA is designed to be general to implement the distributed training algorithms of existing systems. It is supported by Apache Software Foundation.

8. Deeplearning4j is the first commercial-grade, open-source, distributed deep-learning library written for Java and Scala. It is designed to be used in business environments, rather than as a research tool.

9. Convnet.js is a Javascript library for training Deep Learning models (mainly Neural Networks) entirely in a browser. No software requirements, no compilers, no installations, no GPUs, no sweat.

10. Torch is a scientific computing framework with wide support for machine learning algorithms. It is easy to use and efficient, fast scripting language, LuaJIT, and an underlying C/CUDA implementation. Torch is based on Lua programming language.

11. Mocha is a Deep Learning framework for Julia, inspired by the C++ framework Caffe. Efficient implementations of general stochastic gradient solvers and common layers in Mocha could be used to train deep / shallow (convolutional) neural networks, with (optional) unsupervised pre-training via (stacked) auto-encoders. Its best feature include Modular architecture, High-level Interface, portability with speed, compatibility and many more.

12. Lush(Lisp Universal Shell) is an object-oriented programming language designed for researchers, experimenters, and engineers interested in large-scale numerical and graphic applications. It comes with rich set of deep learning libraries as a part of machine learning libraries.

13. DNNGraph is a deep neural network model generation DSL in Haskell.

14. Accord.NET is a .NET machine learning framework with deep learning libraries combined with audio and image processing libraries completely written in C#. It is a complete framework for building production-grade computer vision, computer audition, signal processing and statistics applications.

15. darch package can be used for generating neural networks with many layers (deep architectures). Training methods includes a pre training with the contrastive divergence method and a fine tuning with common known training algorithms like backpropagation or conjugate gradient.