No offense to the author, but there's a wealth of research (and tools) out there on this subject - robust optimization, stable theory, etc. (not requiring LLMs and ML buzzwords)
IMO it's not sensitivity analysis in the classical sense because the explanations people are looking for are not about slight perturbations of the model, but usually more drastic ones. I found that, for example, the dual of an ILP is completely useless for explanation purposes. Stuff like "shadow prices" are just not relevant for complex formulations that model supply chain dynamics.
If there is useful theory here, I'd love to hear about it, but it's not in any textbooks I've read.
Please share; I'm super interested in ways I can apply these concepts to a related work, and I have already heard from at least individual whom believes MIP explainability is 'impossible'.
Hard not to feel that most of this research has been lost on the modern practitioner? I'd wager even money that a sizeable portion of new entrants to programming are not even aware of optimization techniques. Never mind if they are familiar with them.
Nor is this limited to optimization techniques. State machines are surprisingly ignored by many.
Makes me think this would be a fun post. What are the techniques that are basically ignored by most current discourse?
I was surprised he said this, since it sounds like he’s doing sensitivity analysis, which has been a thing for longer than I’ve been alive.