Perceptrons: The First Trainable Neural Networks

Teaching Computers to Learn, Part 3

Neural networks have become powerful tools for teaching computers to learn. But how did we train the first neural networks? In this video, we explore the first neural network that had the capacity to learn: the perceptron, developed by Frank Rosenblatt in 1957. We explore how the perceptron works and how it's able to learn from data, and we look at how Rosenblatt's perceptron compares to modern perceptrons.