Abstract: Classification is one of the most common machine learning tasks. In this workshop we will discuss the unusual challenges when dealing with hundreds or even many thousands of distinct classes, focusing particularly on text classification. We’ll cover several aspects of the problem, from taxonomy development and data exploration, through to nifty linear algebra and optimisation tricks you can leverage to help your machine learning algorithms cope. We’ll walk through an Jupyter notebook applying this knowledge to a new, never-been-seen highly multiclass NLP dataset (soon to be made publicly available).
This talk is aimed at those with knowledge of basic machine learning and python. Some neural network experience is preferable but not required.
Bio: Mike is a senior machine learning engineer at Evolution AI, working on Evolution AI’s NLP platform. He is probably most widely known in the machine learning community for a popular blog post about his escapades teaching a neural network to freestyle rap (https://bit.ly/2fsePbZ). He has has been working in data science and machine learning for the last 5 years, for the likes of Ocado and Qubit Technology. His primary areas of expertise are in NLP, probabilistic graphical models and recommender systems.