✳flâneur — a map of the web's best reading

‘Leaky Pipelines’ Directory

gwern.net · 2,856 words · saved by 1 readers

⁠“‘Leaky Pipelines’ Directory ”⁠2014 (index selection (breeding), mental energy, order statistics, probability; backlinks⁠, bibliography⁠): Bibliography for directory statistics/order/selection/pipeline, most recent first: 2 ⁠related tags, 26 ⁠annotations, & 6 ⁠links (parent⁠). Many multi-step processes look like ‘leaky pipelines’, where a fractional loss/success happens at every step. Such multiplicative processes can often be modeled as a log-normal distribution (or power law), with counterintuitive implications like skewed output distributions and large final differences from small differences in per-step success rates. The log-normal distribution⁠ (quincunx⁠ ⁠visualization⁠ from Li2010⁠) is a skewed distribution which is the multiplicative counterpart to the normal distribution⁠: where the normal distribution is conceptually applicable to where many independent parts are added, the log-normal is where those parts instead multiply. A common example is latent⁠ variables multiplying t

+ seg); } }); /*****************************************************************************/ /* Returns a URL constructed from either a fully qualified URL string, or an absolute local URL (pathname starting at root), or a relative URL (pathname component replacing part of current URL after last slash), or a hash (URL fragment) only. (The existing URL() constructor only handles fully qualified URL strings.) The optional baseURL argument allows for qualifying non-fully-qualified URL strings relative to a base URL other than the current page location. */ function URLFromString(urlString, baseUR

Explore this link on the map →

related reading