Semi-Log Transformations of Data(半对数变换数据) Suppose a data set is actually following the trend of some hidden exponential function y = a b[size=-1] x If we take the logarithm of both sides of this equation (any logarithm will do) and use the laws of logarithms (see the section on algebraic representations of logarithms), we get  log(y) = log(a) + x log(b) Now consider a new data set showing  Y = log(y)  vs.  x . If we let  A = log(a)  and  B = log(b) , then  Y = A + B x  This is linear. Linear data sets are easy to recognize.

若能收集到该词条相关资料的跟帖者给予奖励，根据收集资料的水平给予10～200不等的论坛币，当然也可以谈谈自己的观点、其应用领域、上传资料、相关软件介绍等等，请大家尽量不要复制一些搜索引擎很容易收集到的资料。

     我每天都会发送不等数量的词条，所以诚邀不同统计背景的各路大侠发表评论。

     欢迎大家多多跟帖！要数量也要质量哟