Webinar 11: Unsupervised Learning

In the real world, not every dataset we work on has a target variable. This kind of data cannot be analysed using supervised learning algorithms. In such scenarios, we need the help of unsupervised learning. One of the most popular types of analysis under unsupervised learning is Clustering. Clustering is used when the goal is to group similar data points in a dataset. In this lecture, we familiarise ourselves with the idea behind clustering using a foundational clustering algorithm called K-means clustering. In the latter part of the lecture, we will extend our understanding of clustering by analysing a probabilistic algorithm that performs the same task, the Gaussian mixture model.

Sahan is a lecturer affiliated with the Centre for Artificial Intelligence at University College London (UCL) currently contributing to X5GON and HumaneAI projects. He is the deputy programme director of the AI for Sustainable Development MSc programme at UCL, the first programme of its kind in the world and part of the UNESCO Chair of Artificial Intelligence. His research interests lie in the theme: “Improving AI-enabled systems for lifelong learning”. Before joining UCL, he has worked in several research roles in the industry in cybersecurity and personalised advertising domains where he gained experience in user state modelling in a big data landscape.

Webinar 10: Introduction to solving a real world machine learning problem on the Zindi platform

Join Zindi data scientist and experienced hackathon competitor Stephen Kolesh for a walkthrough of Zindi (https://zindi.africa/signup), a competitive data science platform that will host the IAIO practical challenge. Learn how to sign up, explore our existing challenges, and even make your own submissions to get a headstart on the competition! We will walk you through creating a machine learning notebook and making submissions to get a score on Zindi.

Zindi is an AI and data science community where more than 70,000 data practitioners from all over the world gather to learn, grow their skills and get jobs in data. We help solve pressing challenges with AI and connect organisations with a global data talent pool. For more information please visit https://zindi.africa.

Webinar 3: KERNEL METHODS

Kernel methods enable more powerful non-linear functions to be estimated using linear methods through the use of the so-called kernel trick. The lecture will introduce the principles behind this approach through the examples of kernel ridge regression and kernel perceptron algorithm.

John Shawe-Taylor is a professor at University College London where he holds the UNESCO Chair of Artificial Intelligence. He is also the Director of the International Research Centre on Artificial Intelligence (IRCAI) under the auspices of UNESCO in Ljubljana Slovenia.