峰度系数>0,尖峰(that is, a higher probability than a normally distributed variable of values near the mean)、肥尾(that is, a higher probability than a normally distributed variable of extreme values)
峰度系数<0,低峰( a smaller “peak” around the mean)痩尾
国内课本上写的不对,不是直接导致尖峰扁平,而是导致肥尾痩尾。
肥尾一定尖峰吗? 不一定
尖峰=峰更高吗(more peaked?)不是,尖峰
Distributions with zero kurtosis are called mesokurtic, or mesokurtotic. 例子:正态分布,除了这个,其他分布很少
A distribution with positive kurtosis is called leptokurtic, or leptokurtotic.例子: Laplace distribution and the logistic distribution.
A distribution with negative kurtosis is called platykurtic, or platykurtotic. I例子:均匀分布
这个图肥尾,但是峰值比正态要低

峰度系数是衡量数据分布尖峰程度和尾部厚度的统计指标。大于0表示尖峰且肥尾,小于0则表示低峰且痩尾。常见误解认为峰度直接决定尖峰或扁平,实际上它影响的是尾部极端值出现的概率。正态分布为零峰度的典型例子,而正态分布之外,Laplace和Logistic分布具有正峰度(尖峰肥尾),均匀分布则具有负峰度(低峰痩尾)。
最低0.47元/天 解锁文章
512

被折叠的 条评论
为什么被折叠?



