The Perceptron

Free, in-person workshops at our Old Street campus. Work through a piece of our ML curriculum with our founders. No application required. Overview This workshop is a taster for our machine learning apprenticeship for experienced software engineers and is an opportunity to see if you like this style of learning. It provides a practical introduction to neural networks by starting with the perceptron: one of the earliest trainable models and still the cleanest way to understand how machines learn from examples. We will build a small perceptron that learns to separate labelled points on a 2D grid, then use it to unpack the ideas that still sit underneath modern AI: inputs, weights, bias, activation functions, prediction, error, parameter updates, epochs, and model limits. Background In 1958, Frank Rosenblatt described the perceptron as a system that could learn from experience. Instead of being handed a rule directly, it would see examples, make predictions, receive feedback, and change its internal connection strengths. That pattern is still recognisable in neural networks today. A model receives inputs, combines them with learned parameters, produces an output, compares that output with the expected answer, and updates itself so it performs better next time. The perceptron is deliberately simple. It can learn a straight-line boundary, but it cannot solve problems that need curved, layered, or more abstract representations. That limitation is what makes it such a useful starting point. Once you can see what a single perceptron can and cannot do, the later developments in neural networks make more sense: hidden layers, backpropagation, deep learning, convolutional networks, and eventually transformers. During this workshop, we will focus on the perceptron as the hands-on build. At the end, we will step back and trace what each later breakthrough added, so you leave with a map from the first trainable neural units to the architectures behind modern language models. What we will cover What to bring Please note: this event requires prior experience of software development. Hot-desking You're welcome to hot desk from 10am on the day. ​​​Access information: Please ring the Founders and Coders bell. The space is wheelchair accessible and has a lift. Our toilets are fully accessible. If you have any specific access needs, please let us know when you sign up and we'll do our best to help.