4 ms·
Not quite accurate. "Fewest assumptions" assumes that axioms are equally comparable, but this isn't necessarily true. Obviously one should eliminate redundant a
by naasking 10y ago
Not quite accurate. "Fewest assumptions" assumes that axioms are equally comparable, but this isn't necessarily true. Obviously one should eliminate redundant assumptions, ie. ones that have no effect on observable predictions between two theories, but this provides little guidance for selecting between two theories with drastically different axioms that differ in only a small set of predictions for which we have no data.
Ockham's razor, when formalized as in Solomonoff Induction, suggests preferring theories with the lowest Kolmogorov complexity.