📘 Poisson Distribution – From Scratch (For Notes)
—————
💡 What is Poisson Distribution?
Poisson distribution models the number of times an event occurs in a fixed interval
of time or space, given that:
• Events are independent
• The average rate (λ) is constant
• Two events can’t occur at the exact same instant
—————
📌 Used When:
You’re counting how many times something happens in a fixed time, area, volume, etc.
Examples: - Number of emails per hour - Typos in a book - Accidents at a junction in a
month
—————
📐 Poisson PMF Formula:
[ P(X = k) = rac{^k e^{-}}{k!} ]
Where: - ( X ): number of events - ( k ): actual observed count (0, 1, 2, …) - ( ): average
rate (expected number) - ( e ): Euler’s constant ≈ 2.718
—————
🧮 Derivation from Binomial Distribution
—————
✅ Step 1: Start with the Binomial Formula
[ P(X = k) = inom{n}{k} p^k (1 - p)^{n - k} ]
Let: - ( n o ) (large number of trials) - ( p o 0 ) (small probability) - Keep ( = np ) fixed
(mean)
Substitute:
[ p = rac{}{n} ]
—————
✅ Step 2: Substitute into Binomial
[ P(X = k) = inom{n}{k} (rac{}{n}ight)^k (1 - rac{}{n}ight)^{n-k} ]
—————
✅ Step 3: Approximate for Large n
1. Binomial coefficient
[ inom{n}{k} pprox rac{n^k}{k!} ]
2. Power of probability
[ (rac{}{n}ight)^k = rac{k}{nk} ]
3. Limit of the exponential
[ (1 - rac{}{n}ight)^n o e^{-} ]
—————
✅ Step 4: Combine All
[ P(X = k) pprox rac{n^k}{k!} rac{k}{nk} e^{-} = rac{^k e^{-}}{k!} ]
—————
🎉 Final Poisson Formula
[ P(X = k) = rac{^k e^{-}}{k!} ]
—————
🔍 Interpretation:
Term Meaning
( ^k ) Likelihood of k events
( e^{-} ) Probability of zero events
( k! ) Normalization factor
—————
🧠 Key Properties:
• Mean = ( )
• Variance = ( )
• Skewness = ( 1/ )
• Distribution gets symmetric as ( ) increases