Machine learning (ML) is rapidly advancing and the impact it has on everyday life is becoming increasingly evident, from personalized recommendations and voice assistants to healthcare diagnostics and autonomous driving. One aspect of machine learning that does not get enough attention, however, is the theory behind it. Foundational concepts, such as computational geometry and data distribution, play a crucial role in enabling these technologies to work effectively and reliably. These theories provide the backbone for ML models to recognize patterns, process complex data, and adapt to new information, allowing ML applications in diverse, real-world settings.

In this talk, we will present some foundational theories that support machine learning, focusing on computational geometry and data distribution. These theories, though often overlooked, are essential to the accuracy, efficiency, and adaptability of machine learning systems. We’ll provide specific examples to show how these approaches and ideas are applicable to help us to solve machine-learning problems.