I tried to do the error accuracy thing and worked on it hard but... I wasn't able to prove the principle and I'm not sure what I did wrong but I don't get an accuracy increase of 1/sqrt(N). Maybe you can tell me what I'm doing wrong because I'm feeling stymied by this.

Discussed in class. I wrote up a version that has the beginnings of some of the same ideas, showing that the standard errors (that is, the variation in the results) decrease as the number of times you do the experiment goes up.

http://cs.marlboro.edu/ courses/ fall2014/jims_tutorials/ gschacht/ nov_20
last modified Thursday November 20 2014 2:49 pm EST

name last modified size

---

estimate_accuracy.ipynb Nov 20 2014 9:16 am 14.5kB jims_version.ipynb Nov 20 2014 2:49 pm 13.0kB