CNN and RNN are amongst most important algorithm of Neural Network family, also they differ in their network process and solving problems.<\/p>\n<\/p>\n
So talking about their differences:<\/p>\n<\/p>\n
CNN<\/strong> are used to solve classification and regression problems<\/strong> and RNN<\/strong> are used to solve sequence information<\/strong>.<\/span><\/p>\n<\/p>\n CNN are used for 2D image data, RNN are used for sequenced data(example time stamped sales data, sequence of text, heart beat data, etc).<\/span><\/p>\n<\/p>\n So does “Sequences” means ?<\/span><\/p>\n Example of Sequences:<\/span><\/p>\n<\/p>\n 1) Time Series Data<\/span><\/p>\n 2) Sentences<\/span><\/p>\n 3) Audio<\/span><\/p>\n 4) Car Trajectories<\/span><\/p>\n 5) Music<\/span><\/p>\n<\/p>\n A human would easily predict the future of sequence [1,2,3,4,5,6], and the most obvious answer here is 7. Same thing we do with Recurrent neural network. Predicting the future by building a neural network that can learn from the history.<\/strong><\/span><\/p>\n<\/p>\n So to do this we tell neuron about the previous history of output<\/strong> that came out of the neural network and to do that we simply feed the output back again as an input.<\/strong><\/span><\/p>\n<\/p>\n Take the image as a visual explanation<\/span><\/p>\n A simple neural network looks like this:<\/span><\/p>\n But a Recurrent neural network works like this<\/span><\/p>\n<\/p>\n So a full Recurrent neural network works like this.<\/p>\n Cells that are function of inputs from previous time steps are also known as memory cells.<\/span><\/p>\n<\/p>\n RNN are also flexible in their inputs and outputs, for both sequences and single vector values.<\/span><\/p>\n Artificial Neural Network with three neurons.<\/span><\/p>\n Recurrent Neural Network with three neurons<\/span><\/p>\n How RNN are flexible in their inputs and outputs?<\/span><\/p>\n<\/p>\n <\/p>\n Major Disadvantage of RNN<\/span><\/p>\n<\/p>\n 1) We only remember the previous output, it would be great if we can track the longer history.<\/span><\/p>\n 2) Another major issue is “Vanishing Gradient”.<\/span><\/p>\n<\/p>\n Which I will be talking about in my next blog.<\/p>\n Comment below what you think.<\/p>\n<\/p>\n<\/div>\n Go to Source<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<\/a><\/span><\/p>\n
<\/a><\/p>\n
<\/a><\/p>\n
<\/a><\/p>\n
<\/a><\/span><\/p>\n
<\/a><\/p>\n<\/p>\n
<\/a><\/p>\n
<\/a><\/p>\n
<\/a><\/p>\n