8. Partial Differential Equations

A CDF is a function that maps a real number 𝑥 to the probability that a random variable 𝑋 will take a value less than or equal to 𝑥:

𝐹(𝑥)= 𝑃(𝑋 ≤ 𝑥)

Key Points: • 𝐹(𝑥) ∈ [0,1] • Non-decreasing: As 𝑥 increases, 𝐹(𝑥) does not decrease. • Right-continuous: lim 𝐹(𝑥) = 𝐹(𝑎) 𝑥→𝑎+ • lim 𝐹(𝑥) = 0 and lim 𝐹(𝑥)= 1 𝑥→−∞ 𝑥→∞

The Cumulative Distribution Function (CDF) is a mathematical tool used to describe the probability properties of random variables. It essentially tells us how likely it is that a random variable will fall below or reach a certain value, denoted as 𝑥. The formula 𝐹(𝑥) = 𝑃(𝑋 ≤ 𝑥) specifies this relationship. Important characteristics of CDF include:

1. The values of 𝐹(𝑥) range from 0 to 1.
2. It is non-decreasing, meaning as we move to larger values of 𝑥, the probability will not decrease.
3. At extreme limits, as we approach negative infinity, the CDF approaches 0, and at positive infinity, it approaches 1.

Imagine you're waiting for a bus at a bus stop. The CDF can be likened to the bus schedule that tells you what the probability is that a bus will arrive by a certain time. If the bus usually arrives between 5 and 10 minutes, then after 5 minutes there's a 25% chance (𝐹(5)) the bus has arrived, and by 10 minutes, there's a 100% chance (𝐹(10)) – it has already arrived.

If 𝑋 is a discrete random variable with probability mass function (PMF) 𝑝(𝑥), the CDF is:

𝐹(𝑥) = ∑𝑝(𝑡) 𝑡≤𝑥

Example: Let 𝑋 represent the result of rolling a fair six-sided die. • 𝑝(𝑥) = 1/6 for 𝑥 = 1,2,3,4,5,6 • 𝐹(3) = 𝑃(𝑋≤ 3) = 𝑝(1)+ 𝑝(2)+ 𝑝(3) = 1/6 + 1/6 + 1/6 = 0.5

For discrete random variables, such as the outcome of rolling a die, the CDF is calculated by summing the probabilities of all possible outcomes that are less than or equal to a given value 𝑥. In the case of a six-sided fair die, each outcome (from 1 to 6) has an equal probability of 1/6. To find the CDF at 𝑥 = 3 (i.e., the probability that the result is less than or equal to 3), we sum the probabilities of rolling a 1, 2, or 3, which totals 0.5.

Consider a simple game where you score points by rolling a die. Knowing the CDF helps you understand your chances of scoring at most 3 points in one go. If you think of the game as a race, the CDF represents the finish line: it tells you how many 'runners' (points) have crossed it by that roll.

If 𝑋 is a continuous random variable with probability density function (PDF) 𝑓(𝑥), the CDF is:

𝐹(𝑥) = ∫ 𝑓(𝑡) 𝑑𝑡 −∞

Example: If 𝑓(𝑥)= 2𝑥 for 𝑥 ∈ [0,1], then:

𝐹(𝑥) = ∫ 2𝑡 𝑑𝑡 = [𝑡²]𝑥 = 𝑥² 0 So, 𝐹(0.5) = (0.5)² = 0.25.

For continuous random variables, the CDF is found by integrating the probability density function (PDF) up to a certain point 𝑥. The integral accumulates probabilities from negative infinity to 𝑥, providing a total probability of all values less than or equal to 𝑥. For example, if the PDF is defined as 𝑓(𝑥) = 2𝑥 for values between 0 and 1, the corresponding CDF results in expressions derived through integration, demonstrating how probabilities build up over the range.

Imagine measuring the height of people in a room. If you think of the PDF as a smooth curve showing how many people fall within various height ranges, the CDF would be like keeping a tally of how many people are shorter than a certain height. If that height is 150 cm, the CDF tells you the cumulative proportion of people below that height, which helps to understand demographics, just like how the cumulative experience of life shapes who we become.

1. Monotonicity: 𝐹(𝑥) is a non-decreasing function.
2. Limits: o lim 𝐹(𝑥) = 0 o lim 𝐹(𝑥) = 1
3. Continuity: o Discrete RV: Jump discontinuities o Continuous RV: Continuous function
4. Right-continuous: Ensures integration in PDEs is well-defined.
5. Differentiability: If 𝐹(𝑥) is differentiable, then 𝑓(𝑥) = 𝑑𝐹(𝑥)/𝑑𝑥.

The properties of the CDF provide critical insights into the behavior of probabilities. First, monotonicity implies that the CDF never decreases; it either remains constant or increases as 𝑥 increases. Additionally, the limits show that as we look towards very negative numbers, the probability starts at 0, and as we reach very positive numbers, it approaches 1. For discrete random variables, the CDF may jump at certain points (discontinuities), while for continuous variables, it is a smooth function. Being right-continuous is vital for mathematical integrity, particularly when dealing with integrations in PDEs. Furthermore, if the CDF is differentiable, the derivative gives us the PDF, linking the graphical representation of probabilities to actual numerical behaviors.

Consider the CDF as a staircase where each step represents a different height of people in a group. You can only move up as you exceed each person's height. Sometimes you may skip some heights (discrete), but if you're looking at a ramp (continuous), you can slide up smoothly without jumping. The stairway's structure (properties) helps us understand if anyone is shorter or taller, just as the CDF helps us make sense of probabilities.

For a continuous random variable 𝑋 with PDF 𝑓(𝑥):

𝐹(𝑥) = ∫ 𝑓(𝑡) 𝑑𝑡 −∞

and 𝑓(𝑥)=

d𝐹(𝑥)

d𝑥

This relationship is crucial when solving probabilistic PDEs, where the evolution of a probability density over time (e.g., diffusion, heat conduction) is tracked.

The relationship between the CDF and PDF provides a powerful mathematical connection vital for understanding probability distributions. The CDF is fundamentally the aggregate of probabilities over an interval, represented through integration of the PDF. Conversely, if we want to know how a probability distribution changes at any point, we can find that by differentiating the CDF, yielding the PDF. This relationship is particularly useful in applications like partial differential equations, where probabilities evolve over time, as seen in processes like diffusion or heat conduction.

Think of the CDF as a slow-flowing river that accumulates water (probability) as it flows, while the PDF is the speed of the current at different points. If you want to know how fast the river's current (probability) is at a particular point, you look at how much water has flowed by (CDF) until that point and differentiate it. This understanding is crucial in dynamic scenarios like predicting weather patterns or heat transfer in materials.

1. Heat Transfer Problems with Uncertainty: o Probabilistic boundary conditions can be modeled using CDFs.
2. Reliability Engineering: o CDF used to determine failure probabilities over time.
3. Random Inputs in PDEs: o CDFs model the probabilistic behavior of initial/boundary conditions.
4. Signal Processing: o CDF used in noise analysis and error probability calculation.
5. Stochastic PDEs: o CDF and PDF describe random solutions or source terms in PDEs.

CDFs have numerous applications in various engineering fields, particularly when working with uncertainties. In heat transfer, CDFs can help model how changing conditions affect the heat flow at boundaries. In reliability engineering, CDFs inform engineers about the likelihood of system failures over time, enabling better design and maintenance decisions. For PDEs, which often utilize initial or boundary conditions that can be random, CDFs help frame these uncertainties. In signal processing, CDFs can be applied to analyze noise levels and calculate error probabilities. Lastly, in stochastic PDEs, CDFs and PDFs work together to characterize random variables that influence systems modeled by these equations.

Imagine you're a chef in a busy restaurant, and you must adjust your menu based on various unpredictable factors like food supply and customer demand. Using CDFs is like having a well-refined system that helps you gauge the potential outcomes of daily operations – like estimating the probability of running out of key ingredients. This understanding allows you to make better decisions about what meals to prepare, much like engineers use CDFs to navigate uncertainties in their designs.

In some advanced PDE problems, especially involving random fields or stochastic processes, the solution 𝑢(𝑥,𝑡) may be a random variable at each point in space-time. In such cases, the CDF of 𝑢 helps describe the distribution of outcomes at any given point. For example, in a stochastic heat equation:

∂𝑢/∂𝑡 = 𝛼 ∂²𝑢/∂𝑥² +𝜂(𝑥,𝑡)

where 𝜂(𝑥,𝑡) is a random forcing term, we may study the CDF of 𝑢(𝑥,𝑡) at a fixed point to understand the spread of heat under random conditions.

In complex PDE scenarios, particularly those that involve randomness, the solution can itself be a random variable which prompts the use of CDFs. For instance, in equations modeling heat distribution where there are unpredictable influences (represented as 𝜂(𝑥,𝑡)), the CDF becomes a tool for understanding how the heat distribution varies with the random input. By examining the CDF at a specific point, we can gain insights into the potential variations and behaviors of heat flow, helping predict how the system changes over time under uncertain conditions.

Consider a chef again, this time working with unpredictable ingredient qualities that vary each batch. Each meal you prepare may yield variable results based on these factors – some might be spicier, others sweeter. Using the CDF helps you visualize possible outcomes based on your previous experiences. This way, you're ready for whatever comes your way in the kitchen, much like analyzing outcomes in the face of randomness when solving PDEs.

--