PyData Amsterdam 2017
The increased use of machine learning (both simple and advanced) to make decisions that impact business and at times, our culture and our lives, is simply a fact of our current world. Determining ways to make accountable, fair and ethical decisions when using AI or machine learning is a subject of much debate and current research. Join the conversation and explore how we can help build a more just
This talk aims to cover the current research in training and producing ethical and fair machine learning models. Based on many papers released in the past year, we can see we have a massive problem when it comes to AI and machine learning algorithms and models which make (often accurate) biased decisions based on race, location and gender. As machine learning researchers, data engineers and scientists, our decisions in choosing accuracy over ethics (or the opposite) can impact our lives, our research but also the world. This talk will provide questions and (some) answers about how to fight prejudicial training and work to create less prejudiced models in a biased world.