The neural networks in the human brain serve as the basis for ANNs, which are designed especially for machine algorithms in order to mimic the human brain’s processing ability. The basis of Deep Learning is neural networks. For those who don’t know about neural networks, let me tell you that they are responsible for carrying information throughout the brain and allowing communication to occur between them through neurons within milliseconds.
During these studies, scientists were astonished at how neurons work and how quickly data can flow from neurons to the brain as it produces results in a short period of time. Researchers studied the neuron and its function from the beginning since it was responsible for transmitting information through electrical impulses and chemical reactions. This leads to the idea of mimicking this process in machines to maximize performance and save time.
I will demonstrate how a single neuron functions and then communicates with other neurons so that the input data can be sent to the brain with the help of the above diagram.
Data is collected from sensory organs by dendrites, which initiate chemical reactions that are then sent to the nucleus. The data from the nucleus is transmitted through the axon in the form of electrical impulses to the axon terminals. Essentially, the axon terminal functions as an input layer for the next neuron, but there is a gap between two neurons called a synapse that prevents the neurons from connecting. In answer to the question, how does the data get transferred further, the electrical impulses trigger chemical reactions which in turn enable the data to flow through the synapsis gaps to the dendrites of the next neuron which then converts the same chemical reaction into electrical impulses to carry out the further process. From here the brain processes the data and makes a decision as an input to neurons, which will then transfer the data to the targeted cell for action based on the forward and backward propagation of the neural network.
One can observe the collective work of billions of neurons in order to complete the task eventually inspired the artificial neural network to work in similar fashion.
Currently, the human brain is the most complex system in terms of processing for desired output, such as decision-making. It can perform complex tasks such as image recognition and data analysis within a short period of time. ANN is far more advanced than conventional digital computers because it can handle nonlinear tasks and works parrelally. As a result of its massive structure of networks, it is able to learn and generalize knowledge. ANNs take the input data and process it using suitable algorithms to produce the desired output which is then used to test the accuracy of prediction on another set of data.
In the diagram as you can see that both inputs and output has been provided to the model and ANN is being used to carry out the function.
The input layer consists of x1-xn, which are inputs. These Inputs are fed to the ANN through hidden layers, and there can be more than one. Wij is the weight assigned to each input, or synapsis. The weights and input are then multiplied together and passed on as inputs to the hidden layers, which also use forward propagation until the output y1-yt is generated.
In some cases When the output is not desired as fed, the network follows back propagation methods with error signals to hidden layers based on actual output and continues this process of forward and backward propagation until it predicts the desired output with idle accuracy. This process can take days and even months but it is worth the time as with the help of neurons, ANN processes learning algorithms that learn from their environment and past experiences once trained.
A wide range of mediums are using this ability of machines to solve such complex problems in order to accomplish specific tasks. Here are some examples, Image processing, Weather forecasting, Music Composition and many more. ANNs are performing operations today that were unimaginable a few years ago. This illustrates how unpredictable ANNs can be in the future, but one thing is sure, the future applications for ANNs will be spectacular and will simplify much more complex processes in a way that will have an immense impact on time and performance.
This concludes this topic and I hope you find something useful and helpful and that you learn something new. Thank you for your time.