Optimization Concepts for Data Scientists
Part of the Data Science Go (DSGO) Virtual Conference, October 15, 2021
Outline:
- What is optimization?
- Why is optimization important to data scientists?
- How simulation is used to validate and select an optimization model (using an example of locating trauma centers in Korea)
- Shadow prices may be used as marginal estimates of opportunity costs, and why this is important (using an example of matching advertisements to viewers)
Presentation on Data-Driven Optimization
Part of the UCI Data Science Initiative, October 24, 2014
Key Takeaways:
- Prediction and optimization work well together
- We suggest predicting and optimizing in a joint manner, rather than the more typical approach of predicting first and then optimizing
Notes:
- My part is only until 13:30…if you skip too far ahead, that isn’t really me.
- Strangely, the math wasn’t rendered properly when this video was multiplexed; fortunately, it was properly rendered on-screen during the live presentation. You should still get the big picture of what I am talking about; the details can be found in the corresponding papers.