DahShu's virtual journal club is held monthly to promote the most cutting edge research in the fields of data sciences. Click to download presentation slides and seminar recording materials for each event.
Rui (Sammi) Tang
Fri, Dec 1, 2023, 09:00 – 10:00AM (PDT), 12:00 – 1:00PM (EST)
In this talk we will focus on machine learning interpretability. Modern machine learning has brought many new algorithms and highly accurate models. However, the complexity of these models often obscures their decision-making process, making their predictions and decisions difficult to explain. The lack of interpretability limits the use of these high-accuracy models in many regulated industries. Thus, there is a pressing need in the industry to make these complex models and their predictions interpretable. In this talk we will introduce the concept of machine learning interpretability, its uses cases, and discuss the impact it has in many industries. The talk will also discuss the different dimensions of interpretability and introduce state-of-the-art model-agnostic interpretation methods. We will apply the interpretation methods in several case studies using SAS and show how they can be used to guide and assist both the model-building and decision-making process.
Xin Hunt is a Principal Machine Learning Developer at SAS. She has been working on machine learning research and development at SAS since 2017. Her work focuses on machine learning modeling, model interpretability, and AI fairness. Xin received her PhD in electrical and computer engineering from Duke University in 2017.