Using Decision Model & Notation (DMN) to integrate ML and automated Decision Making. Better Integration and Interpretability of Machine Learning in Automated Business Decisions using DMN.
This presentation describes, with practical examples, how DMN can be used to integrate machine learning into automated business decisions. It shows how to improve the interpretability and post-hoc explainability of ‘black-box’ machine learning models. We also describe patterns of decision models that can be used to implement ensemble machine learning for improved safety, reliability and accountability. The presentation will start with an overview of DMN before considering how deep and shallow machine learning models can be directly integrated into decision models to support both batch and on-line learning. Learn:
- The role of DMN in defining, refining and managing business logic
- How decision modelling improves the probability of successful ML deployment
- Effective patterns for integrating machine learning and AI into automated business decision
- How to use decision modelling to provide explanations for inscrutable machine learning algorithms
- How to supplement machine learning models with heuristics for safety and accountability.