python机器学习-乳腺癌细胞挖掘(博主亲自录制视频)
https://study.163.com/course/introduction.htm?courseId=1005269003&utm_campaign=commission&utm_source=cp-400000000398149&utm_medium=share
对数正态分布(logarithmic normal distribution)是指一个随机变量的对数服从正态分布,则该随机变量服从对数正态分布。对数正态分布从短期来看,与正态分布非常接近。但长期来看,对数正态分布向上分布的数值更多一些。
性质
应用:股票
对数正态分布(logarithmic normal distribution):一个随机变量的对数服从正态分布,则该随机变量服从对数正态分布。
在分析测试中,特别是在痕量分析中,在不少情况下,测定值不遵循正态分布,而是遵循对数正态分布。
在概率论与统计学中,对数正态分布是对数为正态分布的任意随机变量的概率分布。如果 X 是服从正态分布的随机变量,则 exp(X) 服从对数正态分布;同样,如果 Y 服从对数正态分布,则 ln(Y) 服从正态分布。 如果一个变量可以看作是许多很小独立因子的乘积,则这个变量可以看作是对数正态分布。一个典型的例子是股票投资的长期收益率,它可以看作是每天收益率的乘积。
Some common distributions which are not directly related to the normal distribution
are described briefly in the following:
• Lognormal distribution: A normal distribution, plotted on an exponential scale.
A logarithmic transformation of the data is often used to convert a strongly
skewed distribution into a normal one.
Normal distributions are the easiest ones to work with. In some circumstances a set
of data with a positively skewed distribution can be transformed into a symmetric,
normal distribution by taking logarithms. Taking logs of data with a skewed
distribution will often give a distribution that is near to normal