Abstract:

Machine Learning (ML) is a very hot topic today with all of the different AI tools that have reached the market the latest year. The history and development has, however, been a long a winding road. In this talk we are going to explore some of the most used methods in classification (and detection) tasks used before 2012. What happened in 2012? Well, the figurative asteroid, and the literal working Convolutional Neural Networks (CNNs), came to see the light. This year has thus become some sort of rebirth of ML, from a mathematical curiosity, to the hype machine it is today. We will see what methods were used back then and perhaps see how these paved the way for the modern neural networks we see today.