泊松分布公式
概率是数学的基础之一。泊松分布是一种离散概率分布,它表示给定数量的事件在固定时间间隔内发生的概率。当您知道事件发生的频率时,它有助于预测某些事件发生的概率。泊松分布公式很简单,借助这个公式,可以很容易地解决问题。
泊松分布
泊松分布是一种概率分布,用于显示事件在特定时期内发生的次数。它是给定时间段内发生的事件数量的离散概率分布,给定事件在该时间段内发生的平均次数。它是与极其罕见但具有大量独立发生机会的事件的概率相关的分布。
泊松分布公式
假设 X 是一个离散随机变量,可以假设值 0、1、2……那么,X 的概率函数由泊松分布给出。
其中,λ是正常数
推导
This distribution can be derived as the limiting case of binomial distribution by making n very large and p very small, keeping np (=λ) fixed.
The probability of x successes in a binomial distribution is:
P(x) =
Where p is the probability of success, q is the probability of failure, n is the number of trails
P(x) =
As n ⇢ ∞, p ⇢ 0. np=λ
P(x) =
P(x) =
泊松分布图
泊松分布性质
- 泊松分布只有一个参数“λ”
- λ = np
- 平均值 = λ ,方差 = λ ,标准偏差 = √λ。
- 偏度 = 1/λ
- 峰度 = 3 + 1/λ
- 泊松分布呈正偏态且呈尖峰态。
示例问题
问题 1:如果工厂生产的总产品中有 4% 有缺陷。求 50 件商品样本中缺陷品少于 2 件的概率。
解决方案:
Here we have, n = 50, p = (4/100) = 0.04, q = (1-p) = 0.96, λ = 2
Using Poisson’s Distribution,
P(X = 0) = = 1/e2 = 0.13534
P(X = 1) = = 2/e2 = 0.27068
Hence the probability that less than 2 items are defective in sample of 50 items is given by:
P( X > 2 ) = P( X = 0 ) + P( X = 1 ) = 0.13534 + 0.27068 = 0.40602
问题 2:如果药物不良反应的概率为 0.002,确定 1000 个人中超过 3 人会出现药物不良反应的概率。
解决方案:
Here we have, n = 1000, p = 0.002, λ = np = 2
X = Number of person suffer a bad reaction
Using Poisson’s Distribution
P(X > 3) = 1 – {P(X = 0) + P(X = 1) + P(X = 2) + P(X = 3)}
P(X = 0) = = 1/e2
P(X = 1) = = 2/e2
P(X = 2) = = 2/e2
P(X = 3) = = 4/3e2
P(X > 3) = 1 – [19/3e2] = 1 – 0.85712 = 0.1428
问题3:如果一个工厂制造的螺丝总数中有1%是有缺陷的。在 100 个螺钉的样本中找出少于 3 个螺钉有缺陷的概率。
解决方案:
Here we have, n = 100, p = 0.01, λ = np = 1
X = Number of defective screws
Using Poisson’s Distribution
P(X < 3) = P(X = 0) + P(X = 1) + P(X = 2)
P(X = 0) = = 1/e
P(X = 1) = =1/e
P(X = 2) = =1/2e
P(X < 3) = 1/e + 1/e +1/2e
= 2.5/e = 0.919698
问题 4:如果在一个行业中,有 5% 的员工可能会受到电晕的影响。在一组 20 名员工中,超过 3 名员工遭受电晕的概率是多少?
解决方案:
Here we have, n = 20, p = 0.05, λ = np = 1
X = Number of employees who will suffer corona
Using Poisson’s Distribution
P(X > 3) = 1-[P(X = 0) + P(X = 1) + P(X = 2) + P(X = 3)]
P(X = 0) = = 1/e
P(X = 1) = = 1/e
P(X = 2) = =1/2e
P(X = 3) = =1/6e
P(X > 3) = 1 – [1/e + 1/e + 1/2e + 1/6e]
= 1 – [ 8/3e] = 0.018988
问题 5:制造商知道他制造的灯泡中有 2% 的灯泡有缺陷。如果他制造 200 个灯泡,那么少于 4 个灯泡有缺陷的概率是多少。
解决方案:
Here we have, n = 200, p = 0.02, λ = np = 4
X = Number of bulbs are defective
Using Poisson’s Distribution
P(X < 4) = P(X = 0) + P(X = 1) + P(X = 2) + P(X = 3)
P(X = 0) = = 1/e4
P(X = 1) = = 4/e4
P(X = 2) = = 8/e4
P(X = 3) = = 32/3e4
P(X < 4)= 1/e4 + 4/e4 + 8/e4 + 32/3e4 = 71/3e4
= 0.43347
问题 6:在一所大学,某位成员在任何一天缺勤的概率为 0.001。如果有 800 名员工,计算任何一天缺勤人数为 4 的概率。
解决方案:
Here we have, n = 200, p = 0.02, λ = np = 4
X = Number of members of staff absent on any day
Using Poisson’s Distribution
P(X = 4) =
= 0.00767