How do we build machine learning models? How can designers participate in model design?

Machine Learning models are designed to optimize for model performance. Although models can be great at answering the specific questions they are trained to focus on, a model might not be the best way of solving a business need or challenge.

In order to prioritise getting a high accuracy score, sometimes models do not encompass certain attributes. In this way, models themselves do not always reflect the ‘reality’ of the true business problem. Models can struggle to encompass human behavior due to either data limitations or an inability to quantify relational dynamics.

In this interactive, participatory workshop, together with Lauren you will invite friction – including divergent thinking – into the model development process. You will work through a Machine Learning business challenge together and explore interdisciplinary methods for designing models. You’ll learn about model setup and where design can provide a layer of powerful insights to define the Machine Learning use-case.

By being design-led in the Machine Learning approach, more models can be developed that focus on solving true business needs with a good, people-centered user experience.