Some people claim that “when you reduce the level of abstraction, you lose accuracy,” but that’s a low-resolution perception. - All models are wrong. - Abstract descriptions of real events have less information from actual real events. - So accuracy is decreasing. - Increasing abstraction decreases accuracy.
- On the other hand, mathematical concepts are abstract in nature.
- To illustrate with a concrete example, the “generality” that the original model had is lost.
- The characteristic that a model can explain the behavior of a wide range of objects.
- That’s because more information leads to more restrictions.
- To illustrate with a concrete example, the “generality” that the original model had is lost.
- Some people confuse these two phenomena, even though they are two different things.
2019-05-15 First published. Linked from Landing abstract concepts on concrete facts by AI on 2023-09-08
This page is auto-translated from /nishio/抽象度を落とすと正確性が失われるのか using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I’m very happy to spread my thought to non-Japanese readers.