Cover Image for Weighting for Godot
Cover Image for Weighting for Godot
The place where perceptive minds collide. Conservation of energy, optional.
Registration
Welcome! To join the event, please register below.
About Event

​About Event

​Note: we are accepting registrations on a first-come, first-served basis due to lecture theatre capacity. If you attended previous sessions, you will be prioritised. 📍 Room: ELG01, Drysdale Building.

​UPDATE:
We're delighted to welcome guest speaker Bruno Hawkins (MPhil Cantab), who will deliver the lecture in the second half to share his research at UCL and Cambridge, passion projects, and take your questions in a Q&A.

​You can find out more about Bruno here. Or see him mog on Insta here. Please bring thoughtful questions, your laptops, and your curiosity. We'll bring the pizza, cool ML content, and more!

​🎉 LECTURE 4: WEIGHTING FOR GODOT

​The Machine Learning (ML)/AI Lecture Series, led by Martin He, continues! 🚀 Last time we mapped out the ML landscape and built our maths and Python foundations. Now it's time to open up the black box and see how neural networks actually learn.

​Whether you come from STEM, Medicine, Business, Finance, Linguistics, Law or the Humanities, this session will take you from "what is a neural network?" to understanding the techniques practitioners use to train them well. Missed an earlier session? No problem, you're still very welcome to join! Just catch up using the link.

​📚 What to Expect:

  • ​🧠 Neural Network Fundamentals: See how a network makes predictions with the forward pass, and how it learns through gradient descent. We'll cover mini-batches, adaptive learning rates, and compare object-oriented and functional approaches to building models in code.

  • ​🎛️ Hyperparameter Tuning: Learn how to choose the settings that make or break a model: epochs, optimisers, batch size, learning rate, dropout rate, number of layers, and activation functions, plus how cross-validation helps you tune them properly.

  • ​🛡️ Regularisation: Discover how to stop your model memorising the training data, using L1 and L2 regularisation, batch normalisation, and data augmentation.

  • ​🌐 Beyond the Basics: A first look at the architectures behind today's AI, including RNNs, CNNs and Transformers, and how they power NLP and Computer Vision.

  • ​🎤 Guest Talk & Q&A with Bruno Hawkins.

  • ​🍕 Please give your pizza preferences when registering!

​Whether you're new to ML, revising the fundamentals, or curious about how models like ChatGPT actually learn, this session will give you the core toolkit for training neural networks.

​🔥 Bring your laptop, your curiosity, and your questions!

​We look forward to seeing you there!

​Useful Links

​WhatsApp:
City Optimise Society: Join WhatsApp Group
City DSAI Society: Join WhatsApp Group

​LinkedIn:
/city-optimise
/city-data-science-and-ai-society


​Instagram:

​cityoptimise

​citydsaisociety

Location
City St George's, University of London
Northampton Square, London EC1V 0HB, UK
The place where perceptive minds collide. Conservation of energy, optional.