WEBVTT 00:02.200 --> 00:08.380 Hello and welcome to this tutorial here we will learn the basic building block of artificial neural 00:08.380 --> 00:09.070 network. 00:09.250 --> 00:10.690 That is the new ROM. 00:12.190 --> 00:16.060 So this is the actual biological structure of a neuron. 00:17.320 --> 00:19.700 This is the main part of a neuron. 00:20.920 --> 00:26.140 The body of neuron has branches and these branches are called Dad's then lives. 00:27.550 --> 00:32.680 And it has a long day, which is called as Exon for a neuron. 00:32.680 --> 00:36.760 The androids are the receivers of Signal and Axon is the transmitter. 00:37.480 --> 00:38.230 No, no, Don. 00:38.380 --> 00:40.810 An important point here to function. 00:40.900 --> 00:41.920 A human brain. 00:42.130 --> 00:45.160 A single neuron can do much to do that. 00:45.250 --> 00:47.090 There are billions of neurons. 00:47.530 --> 00:51.580 These all neurons work together, transmit and receive signals. 00:51.820 --> 00:55.210 And as a result, human brain functions properly. 00:55.960 --> 01:01.690 Now the deep learning and artificial intelligence scientists are inspired by the functioning of brain. 01:01.930 --> 01:07.330 And they are trying to mimic the functioning of the brain and to mimic the functioning of a brain. 01:07.540 --> 01:08.980 They are using neurons. 01:12.010 --> 01:13.810 So this is a single neuron. 01:14.440 --> 01:16.030 It is also called as node. 01:17.440 --> 01:24.670 Now this neuron is receiving input signals like daybreaks to receive signals after processing the input 01:24.670 --> 01:25.240 signal. 01:25.390 --> 01:29.290 This neuron is transmitting the output that is output signal. 01:30.070 --> 01:33.400 So here we are trying to mimic a biological neuron. 01:35.550 --> 01:39.480 Now, this single neuron is receiving input signals from other neurons. 01:39.750 --> 01:45.650 These other neurons are denoted as X1, x2 and up to X and. 01:47.000 --> 01:54.810 And the neurons transmit signal to sign UPSs after processing the input signal from other neurons. 01:55.020 --> 01:56.900 This neuron transmits output. 01:57.480 --> 02:01.250 This output signal by this neuron is denoted as VI. 02:02.040 --> 02:05.700 So the single neuron receives the input signal from other neurons. 02:05.910 --> 02:08.280 Then it processes that input signal. 02:08.490 --> 02:12.120 And at the end it transmits that signal as an output. 02:12.840 --> 02:15.360 This is the working of a single neuron. 02:17.630 --> 02:22.240 No distinguish neuron is receiving input signals from other neurons, right? 02:22.860 --> 02:29.240 No, no, don't hear the input neurons that are transmitting signal to a single neuron are independent 02:29.240 --> 02:30.140 variables. 02:30.440 --> 02:34.100 These are the independent variables from a single observation. 02:34.310 --> 02:40.970 You can think it as a role in an Excel sheet, like the information about an employee in a room. 02:41.270 --> 02:44.750 Age, employee I.D., date of work and so on. 02:45.440 --> 02:47.330 This is a single observation. 02:47.480 --> 02:56.870 Remember this independent variable, one independent variable to add up to independent variable X and 02:58.520 --> 03:00.790 after processing the input signals B how? 03:00.860 --> 03:01.730 Output here. 03:01.940 --> 03:07.540 That is why this output can be a continuous value like pride. 03:07.820 --> 03:10.730 All this output can be a binary output. 03:10.880 --> 03:12.430 Like s no. 03:12.800 --> 03:17.360 And other than these two types, the output can be a categorical output. 03:19.460 --> 03:23.600 So this is a signal of their vision that a neuron is processing here. 03:24.140 --> 03:27.800 Throughout the process, we are processing a single observation here. 03:28.010 --> 03:31.880 And at the end, we have output from that single observation. 03:34.210 --> 03:39.400 As we had discussed earlier, neurons transmit and receive signals through design offices. 03:41.310 --> 03:42.530 In deep learning did sign. 03:42.650 --> 03:43.790 These are called as. 03:44.460 --> 03:45.690 As you can see here. 03:45.930 --> 03:48.490 Wait one wait to up to it. 03:48.690 --> 03:55.720 And by adjusting the weights, a neural network, Lon's as of now, just remember that to the base, 03:55.860 --> 03:57.360 a neural network alone. 03:59.490 --> 04:05.100 Let us understand what is happening inside a mood on inside a neutron. 04:05.230 --> 04:08.520 We did some of these all raids are calculated. 04:09.720 --> 04:16.200 And after that activation function is applied here, the activation function calculates debated sum 04:16.260 --> 04:19.540 of input and then it adds the bias. 04:19.860 --> 04:24.570 And after that, it decides whether a neutron should be activated or not. 04:25.320 --> 04:31.260 In simple words, we can say that the activation function decides whether a signal should be passed 04:31.260 --> 04:32.520 further or not. 04:33.240 --> 04:36.450 So this is how a signal is processed by a neuron. 04:36.660 --> 04:41.640 It receives signal calculous, debated some outplays activation function. 04:41.820 --> 04:44.190 And at the end it transmits the output. 04:44.910 --> 04:48.840 So theoretically, this is all about the working of a single neuron. 04:52.050 --> 04:59.100 As you can see here, this is a neural network and this neural network is made up of multiple neurons. 04:59.640 --> 05:02.700 There is an input layer here, then layers. 05:02.850 --> 05:05.760 And the output layer here. 05:05.820 --> 05:12.090 Each neuron is receiving input signal processing that input signal with the help of activation function. 05:12.300 --> 05:14.780 And at the end, transmitting the same. 05:15.480 --> 05:17.760 This is how a neural network works. 05:18.660 --> 05:23.330 So this introductory tutorial about the new long and neural network NCAR. 05:23.640 --> 05:25.410 I will see you in the next one. 05:25.680 --> 05:27.390 buildOn, happy learning.