The power of Causal Inference
At Silverpond we’ve been excited by the power of Causal Inference, and have been actively researching in this space. Whilst it isn’t as popular as deep learning today, we believe that it’s impact could be just as significant.
Last Thursday, Lizzie Silver from our team gave a talk at the Data Science Melbourne meetup on commercial applications of causal inference. The talk was recorded and you can watch it on YouTube here: https://youtu.be/Ze6a1DlrcrM
The talk included three parts:
- What causation is, how to represent it, and how causal models are different from predictive models
- Brief examples where causal inference is relevant in commercial contexts
- Ad targeting
- Medical risk scores
- Recommender systems
- Customer experience
- Search engine optimisation
- Case study using causal inference for algorithmic counterfactual fairness. This is Zaiga’s project, which we covered previously in this blog post.
She did a great talk and it was well received.
Please reach out if you’d like to explore causal inference or apply it to a problem; we are always interested in causal projects!