Recurrent Neural Network (simplified)
THE UNDERSTANDING OF RECURRENT NEURAL NETWORK AND THE RELEVANCE TO THE TECHNOLOGY BASE
In my previous post we looked at what is Artificial intelligence, and the goal of artificial intelligence.
From the previous post you understand the concept of AI , Artificial Intelligence is a way of making a computer, a computer-controlled robot, think intelligently, which is similar to the way t humans think. To create an AI you need to know neural networks.A neural network is a way of mimicking the human brain it is the way to design an artificial brain this concept was gotten from biology, the human brain consists of billions of neurons which are connected by synapses. Now we can now model it to achieve what we want. The type of neural network we will be looking at is the RNN which is Recurrent Neural Network and why RNN.
Recurrent network has brought break through to the world of machine learning. Recurrent Neural Network is used across various application and use cases. In this post you will understand the basic concept of recurrent neural network and the application which it can be applied.
Recurrent neural network is a branch of deep leaning that is used in machine learning.
The best way to to see it is that a recurrent neural network is a type of network that have an input layer, hidden layer, and also an output layer.
The input layer receives the input, the hidden layer activation are applied and then finally receives the output.
WHY RECURRENT NEURAL NETWORK
This Network is used in a model that deals with data that changes with time. An example is image captioning, Music, prdiction etc.
A Recurrent Net can Receive a sequence of value as an input and output another sequence of value as an output or it could be a set of data set. An example is language translation from one language to another.
Note: Recurrent Neural Network is very difficult to train due to the problem of vanishing gradient, this problem can be solved or addressed with the use of Gating i.e (LSTM, GRU etc.)
GPU are mostly used to train RNN instead of CPU.
Dont forget to UPVOTE and RESTEEM
I´m trying to learn about this, I think it´s very interesting!
Thank you very much. Do you want to know data science
Yes, I really want to learn about it. I am just starting, but I feel I like it so much, and that is very good. I think it is very useful.