19. Partial Differential Equations

The Poisson distribution is a discrete probability distribution that gives the probability of a given number of events occurring in a fixed interval of time or space, given the events occur with a known constant mean rate and independently of the time since the last event.

If 𝑋 is a Poisson random variable with mean 𝜆, then the probability mass function (PMF) is given by:

𝑃(𝑋=𝑘) = \frac{𝑒^{−𝜆}𝜆^{𝑘}}{𝑘!}, 𝑘 = 0,1,2,…

Where: • 𝜆 = average number of events in a given interval • 𝑒 ≈ 2.71828 • 𝑘! = factorial of 𝑘

The Poisson distribution helps us understand how likely it is that a certain number of events will occur in a fixed timeframe when these events happen independently. The formula given shows how to calculate this probability using a mean value (𝜆) which is the average number of events we expect to occur. The term 'e' is a mathematical constant that helps in computing the PMF accurately for different values of k, which represents the number of events.

Imagine you're at a bus stop, and on average, 2 buses arrive every 10 minutes. If you want to know the likelihood of exactly 3 buses arriving in the next 10 minutes, you can use the Poisson distribution with 𝜆 = 2. It provides a clear method to calculate such probabilities in everyday scenarios.

1. Mean and Variance: Mean = 𝜆, Variance = 𝜆

2. Additive Property: If 𝑋₁ ∼ Poisson(𝜆₁) and 𝑋₂ ∼ Poisson(𝜆₂) are independent, then: 𝑋₁ + 𝑋₂ ∼ Poisson(𝜆₁ + 𝜆₂)

3. Memoryless Nature: Though primarily a property of the exponential distribution, the Poisson process (from which Poisson distribution originates) also assumes independent and memoryless events.

4. Skewness: Skewness = \frac{1}{√𝜆} This shows that the distribution becomes more symmetric as 𝜆 increases.

The properties of the Poisson distribution highlight its unique characteristics. The mean and variance being equal to λ indicates how concentrated the probabilities are around the average number of events. The additive property illustrates that when two independent Poisson processes are combined, the result is still a Poisson process with the sum of their rates. Lastly, as λ increases, the distribution approaches a more symmetric shape, reducing skewness.

Think of it this way: if you're tracking how many cars pass by your house every hour, the average number of cars (λ) gives you a good idea of what to expect, and knowing that two separate streets with independent traffic have their counts can help you predict total traffic. As traffic patterns stabilize (higher λ), the unpredictability decreases, showing that the distribution is becoming more consistent.

The Poisson distribution is derived as a limiting case of the Binomial distribution when: • Number of trials 𝑛 →∞ • Probability of success 𝑝 → 0 • 𝑛𝑝 = 𝜆 remains constant

Let 𝑋 ∼ Binomial(𝑛,𝑝), then: 𝑃(𝑋 = 𝑘) = \binom{𝑛}{𝑘}𝑝^{𝑘}(1−𝑝)^{𝑛−𝑘}

Taking the limit as 𝑛 → ∞, 𝑝 → 0, such that 𝑛𝑝 = 𝜆, we get: 𝑃(𝑋 = 𝑘) = \frac{𝑛^{𝑘}𝑒^{−𝜆}𝜆^{𝑘}}{𝑘!}.

The derivation shows how the Poisson distribution emerges from the Binomial distribution under specific conditions. As we consider an infinite number of trials with an increasingly smaller probability of success, while keeping the product n*p constant, it simplifies to the well-known Poisson distribution model. This transition helps understand situations where we expect many independent occurrences but with low probability.

Imagine flipping a coin many times and counting how many heads appear. If you flip it an infinite number of times, but the coin is weighted such that it almost never lands on heads (low probability), the number of heads in a given timeframe starts resembling a Poisson distribution. It's like a light being flickered on and off – the more you press the switch, the chance of it being off more often shows consistency in randomness.

1. Poisson's Equation in PDEs: The Poisson equation is a partial differential equation of the form: ∇²𝜙 = 𝑓(𝑥,𝑦,𝑧) It is used in problems involving electrostatics, gravitational fields, and heat conduction, where the source term 𝑓 is often related to a Poisson-distributed phenomenon.

2. Telecommunication: Models the number of phone calls or messages received per unit time.

3. Quality Control: Determines the number of defects in manufactured products.

4. Traffic Flow: Models vehicle arrivals at a traffic intersection.

5. Radiation Physics: Describes the number of radioactive decays in a given time frame.

This section highlights the real-world applications of the Poisson distribution in various fields such as engineering, telecommunications, quality control, and physics. Each application connects the theoretical concepts to practical scenarios where Poisson processes can predict occurrences, enhancing decision-making in those areas. For instance, in electrostatics, the Poisson equation helps in calculating potential fields based on distributed charges.

Consider how call centers manage incoming calls. By using the Poisson distribution, call center managers can predict busy times and how many agents are needed based on average call arrivals. This method makes it easier to provide sufficient service during rush hours, similar to how traffic lights change based on predicted car arrivals, ensuring a smooth flow.

Feature Poisson Binomial Normal Type Discrete Discrete Continuous

Domain 𝑘 = 0,1,2,… 𝑘 = 0,1,2,…,𝑛 −∞ < 𝑥 < ∞

Mean Yes (𝜆) No No

Variance Derived from Binomial (limit case) Bernoulli trials Central Limit Theorem

Symmetry Skewed unless large 𝜆 Approx. symmetric for large 𝑛 Symmetric

In this chunk, we compare the Poisson distribution to the Binomial and Normal distributions. Understanding their differences helps us appreciate when to use each type. For instance, while the Poisson distribution deals solely with counts of events in discrete terms, the Normal distribution is used for continuous variables and tends to be symmetrical, whereas the Poisson can be skewed unless λ is large enough.

Think of different measurements: Poisson is like counting how many stars you can see on a clear night (discrete counts), Binomial is like trying to see how many times a die shows a number after multiple rolls (fixed trials), while Normal distribution resembles measuring the heights of trees in a forest, which spread out and align symmetrically around an average height.

Example 1: If the average number of emails received in an hour is 5, what is the probability that exactly 3 emails are received in a particular hour? Solution: Here, 𝜆 = 5, 𝑘 = 3 𝑃(𝑋 = 3) = \frac{𝑒^{−5}5^{3}}{3!} ≈ 0.1404

Example 2: A manufacturing unit produces a defect on average every 2 meters. Find the probability that there will be no defect in a 4-meter length. Solution: Rate = 1 defect / 2 meters ⇒ 𝜆 = 4/2 = 2 𝑃(𝑋 = 0) = \frac{𝑒^{−2}2^{0}}{0!} = 𝑒^{−2} ≈ 0.1353

These examples illustrate practical applications of the Poisson distribution equations in different contexts. In Example 1, we determine the chance of receiving a certain number of emails based on the average rate provided. In Example 2, we evaluate the likelihood of defects in manufacturing using the average defect rate in a fixed length, demonstrating how the models apply to real-life situations.

Imagine you're always expecting a few friends to text you while you’re busy, and you want to know how many you might get at one time. The first example helps you calculate that—and if you're running a factory and want to ensure your products have no defects, the second example gives you insight on managing quality based on averages.

--