What Is Deep Learning? | Deep Learning Explained

 

what-is-deep-learning-deep-learning-Explained

Ever wondered how Google translates a complete webpage to

 

Different language in a matter of seconds or your phone gallery groups images grounded on their position all this is a product of deep literacy but

 

What exactly is deep literacy deep literacy is a,

 

Subset of machine literacy which in turn is a subset of artificial intelligence,

 

AI be a technique that allows a machine to mimic human

 

Behavior machine learning is a technique

 

To achieve AI through algorithms trained with data and finally deep learning is a

 

Type of machine literacy inspired by the structure of the mortal brain in

 

Terms of deep learning this structure is called an artificial neural network

 

Let's understand deep learning better and how

 

It's different from machine learning say

 

We produce a machine that could separate between tomatoes and cherries

 

If done using machine learning we'd have to tell the Machine the features based on which

 

The two can be discerned these features could be


The size and the type of stem on them with Deep Learning on the other hand the

Features are picked out by the neural network without mortal intervention of course

That kind of independence comes at the cost of having a important higher volume of data to train our machine now

 

Let's dive into the working of neural networks here

 

We've three scholars each of them write down the number nine on a piece of paper specially

 

They do not all write it identically the mortal brain can fluently recognize

 

The digits but what

 

If a computer had to fete them that's

 

Where deep learning comes in here's a neural network

 

Trained to identify handwritten integers each number is present as an image of 28 times 28 pixels now

 

That amounts to a aggregate of 784 pixels neurons

 

The core reality of a neural network is where the information

 

Processing takes place each of the 784pixels is fed to a neuron in

 

The first subcaste of our neural network this forms

 

The input layer on the opposite end we've

 

The affair subcaste with each neuron representing a number with the retired layers being between them

 

The information is trans from one layer to a different over connecting channels

 

Each of those contains a value attached to that and hence is named a weighted Channel all neurons have a

 

Unique number associated with it called bias this bias is added to the weighted sum of inputs reaching the neuron which is

 

Then applied to a function known as the activation function the result of,

 

The activation function determines if,

 

The neuron gets activated every activated neuron passes on information to,

 

The following layers this continues up till

 

The alternate last subcaste the one neuron actuated in the affair subcaste corresponds to the input number,

 

The weights and bias are continuously set to produce,

 

A well-trained network so where is deep learning applied in customer support,

 

When utmost people converse with client support agents the discussion seems so real they do not indeed realize,

 

That it's actually a bot on the other side in medical care neural networks detect cancer cells and analyze MRI images to give detailed results

 

Own- driving buses what feel like wisdom fabrication is now a,

 

Reality Apple Tesla and Nissan are only some of the businesses performing on self-driving cars so deep learning contains a vast scope,

 

But it too faces some limitations.

 

The first as we discussed earlier is data while deep learning is,

 

The most efficient way to deal with unstructured data a neural network requires a massive volume of data to Train

 

Let's assume we always have access to the necessary amount of data processing.

 

This isn't within the capability of every machine and,

 

That brings us to our second limitation computational power training and neural network,

 

Requires graphical processing units which have thousands in fact as compared to CPUs and GPUs are in fact.

 

More precious and eventually we come down to training time deep neural networks,

 

Take hours or indeed months to train the time increases with the quantum of data and number of layers in the network,


So heem to have only scratched

 

The surface in fact horse technology is working on a device for the blind,

 

That uses deep learning with computer vision to describe the world to the users replicating.

 

The human mind at all could also be not just an episode of fantasy or too long.