The Future Of Artificial Intelligence And 

Neural Networks

By Hamish Leahy

Thesis

  2023

The foundation of neural network research. It all started in 1957 with Frank RosenBlatt, he was inspired by the biological structures of neurons, the basic preconception of the first neural network made to classify single layer networks able to complete binary classification. This system was created to separate data points and sort them into different categories. Rosenblatt’s interest in the idea of creating the neural network laid the foundation of where we are today. The first artificial neurons, in 1943 Warren McCulloch and walter pitts proposed a computational model for the first simplified mathematical creations of biological neurons. 

This set of artificial neurons were primarily used for receiving inputs and creating predefined inputs that were processed through the neurons, this set the first major milestone in the development of Neural Networks. These artificial neurons provided the first working demo of how individual neurons could co-exist and work in harmony to create a network of neurons replicating the biological brain. 

Then Rosenblatt created the first ever ‘perception learning algorithm’ to train the perception of the first neurons he created the perception learning algorithm. This algorithm was used to adjust the weights ‘the processes’ that the neurons completed to assist in letting them, work in harmony. 

The limitations of early neural networks, the first ever neural networks, while they may have shown major promise for the future, came with major limitations. The first ever neural networks only had a single layer perceptron so it could only process data on a surface level, it also lacked the ability to learn from unlabeled data. 

This prevented the first ever widespread adoption of neural networks until they had the ability to learn from general data inputs and not specified systems. 

The influence that early development of neural networks had was undeniable. Despite their early limitations that showed people the power that computers could have, the concept of artificial neurons was just that a concept up until this point. This raised the funding and scientific interest to continue the development of more sophisticated neural networks. Which pioneered the technology available today. 

This ability to create the first neural networks gave scientists and researchers the ability to expand their research and the knowledge of these neural networks.

Download The Full Thesis Here