Four Tools for Causal Inference
Details
The MLAI Meetup is a community for AI researchers and professionals which hosts monthly talks on exciting research. Our format is:
- 6:00 - 6:20: Socializing
- 6:20 - 6:40: Announcements and AI news
- 6:40 - 7:40: Talk(s) and Q&A
- 7:40 - 8:00 Networking
- 8:00: Head to the nearest pub for dinner
Talks:
David Rawlinson: "Causal Wizard: Enabling subject matter experts to understand the effects of change with causal insights from historical data"
Abstract: When considering changes to the operation or maintenance of complex infrastructure, assets or products, many key questions are inherently causal - what would have happened under different investment scenarios? What is the effect of changing a policy or process? Using ordinary ML techniques to predict the outcome of these counterfactual and interventional problems is likely to give misleading results. But Causal Inference techniques can be combined with ML, expert domain knowledge and historical data to more accurately predict the effects of change. This talk will explain how these elements come together and introduce a new tool called [https://CausalWizard.app](https://causalwizard.app/) that makes these techniques easy for everyone to use.
Ross Pearson: BARD - Bayesian Argumentation, Reasoning and Decision Making
(abstract TBA)
BONUS EXTRA TALK:
Steven Mascaro: CAT - the Causal Attribution Tool
(abstract TBA)
Lizzie Silver: Presenting "Py-Tetrad and RPy-Tetrad: A New Python Interface with R Support for Tetrad Causal Search"
Abstract: Causal discovery answers the "what causes what" type of questions, outputting causal graphs that can be used for causal effect estimation. Tetrad software contains a wide array of algorithms for causal discovery, including some particularly fast and scalable implementations. However Tetrad is written in Java, making it difficult to integrate into most stats/ML workflows, as the industry mostly uses Python and academic research tends to use R. Joe Ramsey and Bryan Andrews have developed new Python and R interfaces to Tetrad. I'll give a quick intro to Tetrad, why it's great and where you might use it, and demonstrate how to access it from Python.
-----------------------------------
To apply to speak at a future MLAI meetup, visit https://mlai.melbourne/speak and let us know what you'd like to talk about!
We're also organizing a hackathon: [https://forms.gle/JPPf7YfMeVYNUEtJ6 ](https://forms.gle/JPPf7YfMeVYNUEtJ6)
Image courtesy of Midjourney.