4 ms·
Sorry, but this is an example of bringing in extraneous math and notation to make something well understood seem mysterious. The fun statistics facts being abu
by jmount 14y ago
Sorry, but this is an example of bringing in extraneous math and notation to make something well understood seem mysterious. The fun statistics facts being abused here are: 1) moment summaries (sum 1, sum x, sum x^2 ...) are easy to manage; 2) many important summary statistics (count, mean, variance, kurtosis ...) are easy functions of moments; 3) once you decide to think parametric you can use that many distributions are completely determined by the first one, two or three sufficient statistics. But since all of the above can be clearly taught- it isn't as exciting.
- jmount 14y agoAlso if we want to get all math abusey (using graduate level concepts to do basic work): how about something like "moments are a co-product"? The emphasized property in the article is calculations on the raw data can be done on the moments. So all the calculations of interest map through the moment summaries (hence you can think of them as a co-product). Surely that is more exotic than just saying "monoid." And monoid is kind of used up by the crowd that says "free monoid" when all they mean is "set of all strings."