(T) I wish I could have an application that would help me to explore the causal relationships of various inputs on an outcome, so that I could play between different inputs to explore different outcomes. We are still in an early phase of applications of causal inference from technology to business. Hopefully one day that application would come and help us to make better decisions. Could be quite hopefully for our political leaders when understanding the impact of fossil fuels on our future 😦
In the meantime, enjoy two lectures on causal inference: one focusing on technology, and another focusing on business.
Causal inference in technology:
Following is a talk that I attended this week given by Professor Daniel Malinsky from Columbia University on “explaining the behavior of black-box prediction algorithms with causal learning” at the Online Causal Inference Seminar.
Basically, this is an attempt to explain the choice of features made by a convolutional neural network in order to reach its prediction. Simply said, the explanation is based on using a causal graphical representation e.g. a partial ancestral graph (PAG) that articulates an equivalence between the causal inference model and the deep learning model. Obviously, much more need to be done on that topics!
Causal inference in business:
Following is another talk that I attended last year given by Sean Taylor from Lyft on leveraging causal inference with data science for ride sharing at the 2020 Causal Science Meeting.
Note: The picture above is a combine harvester in Brittany.
Copyright © 2005-2021 by Serge-Paul Carrasco. All rights reserved.
Contact Us: asvinsider at gmail dot com
Categories: Algorithms, Causal Inference, Mathematics