ConvNetCS Library is a Rewrite of http://cs.stanford.edu/people/karpathy/convnetjs/started.html Convnetjs Library in C# and .Net Framework.
Thanks to Andrej Karpathy ConvNetCS Library is an easy to understand Deep Neural Network even without the Math knowledge.
the network can be built from several layers including :
 1-  Input Layer
 2-  Full Connected Layer
 3-  Dropout Layer
 4-  Convolution Layer
 5-  Local Response Normalization
 6-  Max Pool Layer
 7-  Maxout Layer (Activation Layer)
 8-  Relu Layer    (Activation Layer)
 9-  Sigmoid Layer (Activation Layer)
 10- Regression Layer (Loss Layer) 
 11- SVM Layer 	(Loss Layer)
 12- Softmax Layer (Loss Layer)

 
 also the library include the deep q learning implementation along with the same example in the original library.