PyWhy’s mission is to create an open source ecosystem for causal machine learning that promotes the latest technology and makes it available to professionals and researchers. PyWhy will create and host interoperable libraries, tools, and other resources that cover a variety of causal tasks and applications, connected through a common API for fundamental causal functions and focus on the end-to-end analysis process. Most real systems, whether industrial processes, supply chain systems, or distributed computer systems, can be characterized using variables that may or may not be causally related to each other. The evaluation of causal machine learning models and the formalization and integration of field knowledge into machine learning pipelines present significant research problems. Finding the best recognition technique, setting up an appraiser, and performing robustness checks are all steps that are often completed from the beginning as part of the normal process. However, the cases were difficult to understand and validate. DoWhy is one of the existing causality libraries that focuses on various methods of impact assessment, with the general goal of determining the impact of interventions on a variable target. Utilizing the power of graphical causal models, AWS work enhances DoWhy’s current GCM feature set. Judea Pearl, winner of the Turing Prize, created the formal framework known as GCM to model causal relationships between variables in a system. Causal diagrams, which visually depict cause-effect relationships between observed variables with an arrow from cause to effect, are a critical component of GCMs. DoWhy already integrates potential outcomes and graphical causal models, two of the most popular scientific contexts for causal inference, into a single library for outcome assessments. The AWS contribution seeks to strengthen the link between the contexts and the communities of researchers who are committed to them.
