Is the method of teaching the Fold Network even deeper?

Question

In documents like ImageNet Classification with Deep Convolutional Neural Networks

http://www.cs.toronto.edu/~fritz/absps/imagenet.pdf

the teaching method is apparently the main back propagation with stochastic gradient descents.

Despite the fact that CNNs are part of deep neural networks, is it because of the large number of hidden layers? And does this mean that the backprop here belongs to the category of deep learning, because the network is deep, although it does not follow the same scheme as DBN, using the greedy way of learning by word, a real deep learning method?

Thanks for the help and advice.

Answer 1

If you read the Wikipedia Deep Learning page, it says: “Deep Learning is a machine learning industry based on a set of algorithms that try to model high-level abstractions in data using multiple processing layers, with complex structures, or otherwise consists of several non-linear transformations. "

CNN has several levels of non-linear transformation, so it qualifies as a Deep Learning model.

Also in this book from MIT http://www.deeplearningbook.org/ CNN are also part of Deep Learning models.

DBN CNN : - , , DBN .

RNN LSTM, Deep Learning, , backpropagation, backpropagation .

, .

Deep Learning .

Answer 2

1 1 →

2 , 1 , 1 → y = w * x + w0

2 , 1 , →

, .

, CNN CNN Deep Neural Network, , , . , .

• , N 2020 5 1 ,
• , 400x5 - , 2000 - . 5 N-.
• cnn 20x20, , 4x4 5 4x4 - , 80 . , 5 17x17 N-. 1445 N-. , , , .

, ,

• 1 , / , 1 , DBN. CNN, "" " ". , , 1 "backpropogation".

• , . , backprogogate "/ ", . , /. , , backprop .

• : "CNN , DBN". , , DBN .

, . , , , DBN DNN.

(https://www.quora.com/Deep-Learning/What-are-the-difference-and-relevance-between-DNN-and-DBN)

" ( ) , .

, , , , .

- .

Answer 3

deep learning methods are the latest technologies for artificial intelligence in specific convolutional neural networks (CNNs), very effective in pattern recognition, recognition of objects or faces. Many libraries are available for CNN, such as Itorch, theano, Digits, etc. for a deep understanding of neural networks and deep learning click here