What is Deep Learning?
“NEVER THINK THERE IS ANYTHING IMPOSSIBLE FOR THE SOUL. IT IS THE GREATEST HERESY TO THINK SO. IF THERE IS A SIN, THIS IS THE ONLY SIN; TO SAY THAT YOU ARE WEAK, OR OTHERS ARE WEAK”
- By Swami Vivekanand
Is Deep Learning now overtaking the Machine Learning algorithm?
Let us first know what is Machine Learning?
Machine Learning was coined by “ Arthur Samuel” in the year 1959. As we know to perform any Machine Learning algorithm we require a humongous amount of data and very high computation power. But at that time we are not able to generate or store that amount of data, additionally, the computer also doesn’t have enough computation power. So this technology started exhausting, but now this is not the case. Machine Learning is a subfield of Artificial Intelligence.
In the simplest term, we can say that training a machine on a various algorithm using the large dataset to take a decision like a human brain. Machine Learning uses a variety of algorithms but as the dataset complexity increases it’s model accuracy decreases. Hence Deep Learning comes into popularity.
Interesting Facts: One kg weight of DNA require to store World’s entire data.
Comparison between Artificial Intelligence and Machine Learning and Deep Learning as below.
Deep Learning
Geoffrey Hinton was known as Godfather of Deep Learning. He published various papers on deep learning and done various research work in this field. He currently works for GoogleAI. I highly recommend you check out about him and read his research papers. The main concept of building deep learning technology is to mimic how the human brain operates (neural science) because the human brain it seemed to be a most powerful tool for learning, adapting skills and applying skills. If the computer can be copied that it can create wonders.
Famous Deep Learning Networks.
- Artificial Neural Network
- Convolutional Neural Network
Artificial Neural Network
It has an input layer, a hidden layer, and an output layer. As in neuron dendrites act as receiver of signal and axon act as a transmitter of the signal. In deep neural network input layers act as dendrites i.e it takes the various data input, hidden layer act as various other neurons which transmits the data in the network and after various computation (i.e. by applying various activation function) in the hidden layer it produces output at the output layer using backpropagation.
We can say that Artificial Neural Network is the same as the basic machine learning algorithm, but here we add one additional thing, i.e. hidden layers. But because of that hidden layer, accuracy increases to a great extent.
Convolutional Neural Network
It is mostly used for images or video as data inputs. We can define CNN i.e. it is a class of deep neural networks, most commonly applied to analyze visual imagery. It uses minimal preprocessing hence it gives maximum accuracy than Machine Learning algorithms.
Convolutional Neural network consists of the various step before giving inputs layer.
- Convolution operation
After Applying all the above steps, CNN work same as Artificial Neural Network.
A short Video in Which Godfather of AI Geoffrey Hinton gives an overview of the Foundation of Deep Learning. Must Watch.
Originally published at https://knowledgeinai.blogspot.com on January 8, 2019.
