A Brief History of DL
Discover the key milestones in deep learning, including the perceptron, backpropagation, AI winters, and LSTM. Learn how these developments shaped AI and the evolution towards more efficient learning with less data in deep learning models.
We'll cover the following...
Perceptron
In 1943, Warren S. McCulloch and Walter Pitts published a paper, A logical calculus of the ideas immanent in nervous activity, which made a link between mathematics and neurology by creating a computer model based on the neural networks inherent in our own brains based on a combination of algorithms to create a threshold to mimic how we pass information from our own biological network of neurons. Then, in 1958, Frank Rosenblatt published a paper that would be widely considered the ancestor of neural nets, called The Perceptron: A perceiving and recognizing automaton. This was, for all intents and purposes, the first, simplest, and oldest ANN.
Backpropagation
In the 1960s, developments toward backpropagation, or the idea that a model learns from layers of past mistakes as it trains its way through a dataset, made significant strides toward what would ...