问题如下图：
选项：
A. 
B. 
C. 
D. 
解释：
不太理解为什么是成功的概率越大 才会越趋近于正太分布，不是越小吗？
orange品职答疑助手 · 2018年06月05日
泊松分布的参数λ是单位时间(或单位面积)内随机事件的平均发生率，它等于n*p。当n足够大时，泊松分布就可以用正态分布来近似，而此时，λ也因为n的增大而增大，最终趋向于它的最大值也就是1
Ime · 2018年06月05日
有道理 成功的概率应该是指的期望 泊松分布的期望就是np
orange品职答疑助手 · 2018年06月05日
加油:-p
omegayin · 2018年10月12日
为啥lamda的最大值是1啊？
omegayin · 2018年10月12日
泊松分布的期望不是成功的次数么……我查了文献，lamda趋近无穷大时，泊松分布的极限分布才是正态分布。答案有问题，应该选D。http://m.baidu.com/sf_edu_wenku/view/5d9ac306e87101f69e3195e4.html
omegayin · 2018年10月12日
呃，稍微修正一下刚才的结论：答案应该选C没错，是因为n足够大、p接近1的时候，lamda=np一定能够保证足够大，使得泊松分布可用正态分布近似。D选项的问题是，p接近于0的时候，即使n很大，也无法保证lamda=np足够大（例如假设p=1/n的情形，此时即使n很大，也只有lamda=1，泊松分布无法近似为正态分布）。
NO.PZ2016062402000016 When cyou use the normstribution to approximate the Poisson stribution, assuming you have n inpennt trials, eawith a probability of success of p? When the meof the Poisson stribution is very small When the varianof the Poisson stribution is very small When the number of observations is very large anthe success rate is close to 1 When the number of observations is very large anthe success rate is close to 0 The normapproximation to the Poisson improves when the success rate, χ, is very high. Because this is also the meanvariance, answers A anare wrong. In turn, the binominsity is well approximatethe Poisson nsity when np = χ is large. 辅导员，这里也是没明白为啥要接近于1，数据的规模这个是没有问题
When cyou use the normstribution to approximate the Poisson stribution, assuming you have n inpennt trials, eawith a probability of success of p? When the meof the Poisson stribution is very small When the varianof the Poisson stribution is very small When the number of observations is very large anthe success rate is close to 1 When the number of observations is very large anthe success rate is close to 0 The normapproximation to the Poisson improves when the success rate, χ, is very high. Because this is also the meanvariance, answers A anare wrong. In turn, the binominsity is well approximatethe Poisson nsity when np = χ is large. 请问这个success rate为什么要接近1呢 不是np越大就可以越好的approximate么
When cyou use the normstribution to approximate the Poisson stribution, assuming you have n inpennt trials, eawith a probability of success of p? When the meof the Poisson stribution is very small When the varianof the Poisson stribution is very small When the number of observations is very large anthe success rate is close to 1 When the number of observations is very large anthe success rate is close to 0 The normapproximation to the Poisson improves when the success rate, χ, is very high. Because this is also the meanvariance, answers A anare wrong. In turn, the binominsity is well approximatethe Poisson nsity when np = χ is large. P->0, 为什么不选
When cyou use the normstribution to approximate the Poisson stribution, assuming you have n inpennt trials, eawith a probability of success of p? When the meof the Poisson stribution is very small When the varianof the Poisson stribution is very small When the number of observations is very large anthe success rate is close to 1 When the number of observations is very large anthe success rate is close to 0 The normapproximation to the Poisson improves when the success rate, χ, is very high. Because this is also the meanvariance, answers A anare wrong. In turn, the binominsity is well approximatethe Poisson nsity when np = χ is large. 为啥不选possion分布的适用条件不是概率很小，N很大吗？老师讲的
P不可能大，所以np很大的实质就是n很大对吧？