formalizing inductive divergence

[_rck_]

I have been thinking a lot about induction and the larger context of the Peircian triade of in-, de- and abduction lately. For some of my work-related forays into machine learning, but also for my efforts as a historian, the way in which the preconceptions underlying the data influence the potential learning outcomes has been a key element

Comments

[_rck_]

Not sure; I usually tackle these things deterministically, effectively pinning the probabilities at 0 and 1, with the understanding that later one can relax that to any number between 0 and 1.