Announcement_19

I presented at the JSM Annual Meeting in Nashville. We prove that simple, interpretable models can often achieve accuracy comparable to complex ones, which has implications for AI policy.