It is deep learning a broader area mission learning or abroad A subset of mission only mostly bills with your notebooks computations are growing exponentially Year on year the scope AFIA has become bust and is being used in almost all the door It’s in all the verticals and across all the industries Beat idea retail calico B A for safe travel e-commerce Any of this industry the eye And Emily has a future in all the facets of these industries off This video was helpful to you guys If you like this video like share and subscript all channel and stay connected bitters for future it’s.
What neural networks are we’ll start with seeing briefly what a biological neural netbook is Um we won’t go into matter off The details will just do we know of you because it helps to understand what an artificially order network does Um then we’ll see the overview of artificial neural network and a small example Ah pretty small example on Ah what happened and how in your network loans.
We’ll see how we implement Billy and and and different steps that are a part of implementing artificial neural network, Ah one off them being backpropagation than now the weights up station and then the error function Then we’ll also see the somebody off What The overall learning algorithm that the neural network for lows on the Then we’ll see how implemented over the network and are using the same user that same either status that we used to do so that we only know what There’s The us sense of fuse in your love network in art and were able to compare the output of what we did with provisional Carl rhythms and what the neural network gives us Oh okay.
What by logical neural networks you would’ve seen already how They’re the basic building blocks of the um certain What’s that The central knows systems or these other diagrams which may be repetitive for you But then they help us understand what our biological noodle network does not go very deep in tow By Lord T Um So it’s something that you see in the yellow is a new Ron Okay which is similar to a neuron in the artificial neural network Each contains some data about the arm input data variables.
So like the neurons are interconnected at the point called synapses So in the diagram above um you see lead soak away, there is a synapse So neurons are interconnected at the wind closing up this and this is similar to the artificial neural network because of the difference in what way Variables that you take Um that information is passed to the hidden layer And how that information is passed to the hidden layers is through all point with a similar door The synapse in the biological neural network Um two.
Three parts are typically present didn’t by law to communal network Ah one is the dendrite which is the law Um we can stay like terminals off a noodle network with the responsibility to receive signals from the surrounding neurons than ever Accent with Their nothing like a bit so are comparing it to the artificial neural network like there is a part when you go from the input layer to the hidden layer there’s a part in between our which carries the information’s on the accidents similar to that.
It’s obsolete as an abscess are responsible for receiving the signals and learning from the past activities How we correlated to the artificial neural network It’s um because if you know of the concept of back propagation Even the artificial neural network loans from its past activities and to be able to kind of relate the biological neural network pretty artificial neural network to eat off your onshore act as a neuron in the artificial neural network The accident as a part where the information is traversed from one letter off the new network to another layer of neural network and that then right onto Napster’s I used to send and receive information between one neuron to the other And then they also enable the neural network to learn someday passed after with ease And that’s how the AL garden improves itself.
So uh this is essential to I understand Only because the artificial neuron our network gets its name from the biological Your electrical Okay Um, so this is again um smiled I gave them to just show you how the information is passed out with a biological unit Work on a similar kind of thing happened with the arm Artificial unit workers well capital What happens is sold whenever assorted information is um coming toe a neuron when certain information is present in the new Ron Um So there’s this thing called tray shoulder which would also see with the artificial you know left book.
So once that threshold has reached one that special has been achieved The neural network transmits a signal through the Ag zone And the similar kind of thing happened with the artificial neural network is well whenever the threshold is reached our activation function trekkers and then information is passed from the input Um let’s dig up a clear So we’ll see how each of these steps in the biological neural network helps us also understand what this President Lee artificial neural network Okay, so this is Ah how ah basic off how in artifice in your little book work.
The input that you see here is nothing but importantly how variables over the data are that they’re working with Then there are certain weights which are applied ahh to eat of the important media boots Um until the way it’s not applied that nothing but the help the noodle network tell What is the importance of each of the way vehicles So this is also what we saw with the previous ALCAPA TEM’s be alone So each way be able that is present in our row input data set does not have equal importance get each input variable does not have equal in port Instead I’ll be rebuilt with More important they’re very bitter Yes we’re important How that has decided is through these weights So whichever variable is given a higher rate is comparatively more important than a variable Gaza Yes sir Wait So this is how the weights are amused in the northern network.
The set of inputs which are nothing but um the variables from the leaders at that thought of us neurons This is how we correlate an artificial neural network widow biological neural network s So then there read according to the weights that are applied what I just told And finally this is Oh we do kind of some product off the input variables with the weights Some product means to say X one in tow Doubly one place extra int o w topless extra into w tree in so on Okay so the function that you see here are the summation function of the Sigma function Here The skin also be Callas It can’t work function.
This actually performs some product Settle some of the weight on the input that it receives from each of the neurons The next step is to go to the activation function which will decide whether you’re on mill fires the similar to the biological neural network wearing with the threshold we decide whether arm the information should be sent to the next flesh, Okay So if the threshold also whatever some you get using the transfer function are you compared it to the threshold If it exceeds the threshold the activation function decides to fire.