The INFORMS “masses” want…interpretable AI
Many of us think about what the prestige journals want (possibly more than what makes sense). For me, it is a bit new to think about topic selection to increase presentation attendance, even though I have been a cluster chair for three years in Social Media Analytics. (I am warming up to the job.)
Personally, I value exciting applications with real insights and impacts. I value the inspirational integrity of Gary Kaplan’s commitment to lean in medicine (hope it is working). I value taking the emotional temperature of Syria, measuring the impacts of Twitter bots on presidential elections, and running a mass education campaign to improve the US electorate (factspread.org). However, when I look dispassionately at the data on attendance, the rooms are filled with talks about innovative/cutting edge AI methods. This holds even while most of those in the room have never fitted a topic model or used an optimal tree. I guess it is the love of mathematics mixed with the excitement and fear around AI.