Humans have always been fascinated by the ability of science to learn from its surrounding and the technology to mimic things that seem impossible or too common to notice. One of these is the neural networks that are inspired by the human brain. Although there have been immense advancements in this field there still is a part that the technology could not imitate and that is logic. It can predict results based on the inputs but that is mostly based on mathematical logic. And not everything is mathematics. Scientists have now come up with a learning transistor that mimic the brain activity to a better level.
Learning transistors are that part of science and technology that might make it possible to actually mimic the brain functioning in its true sense. As said earlier until now the technology could not establish a new link that is actually related in real life but mathematically cannot be proved. The learning transistor has been successful in establishing this new link among its data. Just like a mother learns that if the baby is eating his fingers he is hungry.
Difference between the new and the old
The normal transistors act as valves that either increase or decrease the output signals. The output mostly depends on the transistors received input. Researchers have developed a new transistor that is the organic electrochemical transistor. The channel in the transistor is the electropolymerized conducting polymer.
The channel of the new transistor has a property of shrinking, growing or completely eliminating itself. This characteristics of the channel help it to manipulate the intensity of the output by becoming more conducting or less. The polymer can also be trained to react to a certain input with a certain output.
Managing the channel
The response or the output of the channel depends mostly on its channel. The growing and shrinking of the channel makes all the difference. When a channel grows that means it increases the degree of polymerization of that material. This thereby increases the number of polymer chains that conduct the signal.
When a channel shrinks it is subjected to high voltage. High voltage overoxidises the material thus making it inactive. These changes can also be obtained by doping and dedoping the material. The changes in the channel can be short term or permanent as per the requirement of the system from the transistor.
Just like the brain
The change in input can manage the strength of the transistor output too. This can create connections that did not exist previously. The gives a synapse-like functioning to the transistor. Synapse is the communication way between two brain cells.
A few technicalities
The software-based Artificial Neural Networks needs to transmit a lot of signals to simulate a single synapse. This task consumes a considerable amount of energy and computing power too. Compared to that the hardware that uses the organic transistor does the same amount of work using single electronic component.
The organic transistor will definitely help in the advancement of the neural network like hardware to work efficiently. Software and hardware can together form a system that can actually mimic the functionalities of the brain.