Explainability is a game-changer in MLOps, enabling organisations to scale the deployment of models whilst minimising risk. In this session the Seldon team will present cutting-edge techniques to better under complex model outcomes.

In many cases, machine learning models can produce accurate predictions, but it can be challenging to understand how and why they made a decision. This is a key blocker for many ML projects, as explanations for ML models are necessary for a number of reasons. Firstly, Data Science teams want to be reassured that models are working, and if the output is wrong they will want to understand why. Lastly, and perhaps most importantly, teams also want to make sure the model is not biased toward making unethical on non-compliant decisions.

Explainable AI, or XAI, is a rapidly expanding field of research that aims to supply methods for understanding model predictions. Here at Seldon, research is at the forefront of everything we do, and in this webinar, our data science team will dive into the latest techniques we’ve released.

Explainability using Seldon can provide a range of insights. The Seldon team will explore the collection of algorithms provided by Alibi Explain and the types of insight they each provide, looking at a broad range of datasets and models and the pros and cons of each. With Seldon, data science teams can justify, explore and enhance their use of ML and minimize risk for models in deployment. The team will introduce the key features of our latest product release, Alibi Explain v0.9.0, including ‘Global Feature Importance’.