39 bookmarks. First posted by rjknight 8 weeks ago.
Why? Because I’ve been sitting in 100,000,000 meetings where people endlessly debate whether the monthly number of widgets is going up or down, or whether widget method X is more productive than widget method Y. For almost any graph, quantifying the uncertainty seems useful, so I started trying. via PocketIFTTT Pocket
15 days ago by gushov
Under some mild assumptions (I’m going to get back to this in a sec and scrutinize it), we can compute the confidence intervals of the mean estimator as: I’ve personally benefitted from memorizing the formula for the confidence interval and think I probably use it more than the previous (Normal based) one. There’s some smart ways you can implement linear regression so that it’s extremely fast, but we’re not going to use those methods because they don’t generalize very well. In fact, minimizing squared loss (which we just did, in the previous snippet) is actually a special case of maximum likelihood! It’s similar to bootstrapping, but MCMC has far better theoretical underpinnings (we are sampling from a “posterior distribution” using Bayes rule), and it’s often orders of magnitude faster.7 weeks ago by sechilds
The hacker's guide to uncertainty estimates It started with a tweet: New years resolution: every plot I make during 2018 will contain uncertainty estimates — Erik Bernhardsson (@fulhack) January 7, 2018 Why? Because I’ve been sitting in 100,000,000 meetings where people endlessly debate whether the monthly number of widgets is going up or down, or whether widget method X is more productive than widget method Y.math python statistics pandas
8 weeks ago by chris.leaman
tags95 article confidence data-science data.science data datamining datasci datascience dataviz data_science dev error errors estimate estimation fehlerrechnung howto ifttt math mathematik nice-thinking numpy pandas plots pocket probability probabiltiy professional_software_development programming py python python3 r science seaborn stat statistics statistik stats toread tut tutorial uncertainty unread uq via:lena visualisation