91Ó°ÊÓ

When dealing with continuous random variables, the concept of a **Probability Density Function (PDF)** is essential. Imagine you have a smooth curve that plots likelihoods across different outcomes—a PDF is this curve. It tells us how probable specific points within a range are, relative to others. For a beta-distributed variable with parameters \(\alpha = 2\) and \(\beta = 2\), the density function is represented as:
\[ f(y) = 6y(1-y), \quad 0 \leq y \leq 1 \]
This formula gives more weight to outcomes around the middle of the possible range (0.5) and less to outcomes near the edges (0 or 1). You can think of it as how much 'density' there is at any spot, much like how we might describe how crowded a place feels.
The key aspects of a PDF to remember are:

• The total area under the curve is always 1, reflecting the certainty that the random variable takes a value within the specified range.
• The PDF itself is not a probability but describes where the probability is relatively more intense.

Understanding the PDF allows us to comprehend the likelihood of a random variable taking on various values.

To ascertain the probability that a random variable is less than or equal to a value, we use the **Cumulative Distribution Function (CDF)**. In simpler terms, the CDF is the sum of probabilities for all outcomes up to a point \(x\). It's like continuously adding the probabilities as you move from left to right along the number line. For a beta distribution with \(\alpha = \beta = 2\), it is expressed as:
\[ F_Y(x) = x^2(3-2x) \]
The CDF of \(Y_{(n)} = \max(Y_1, Y_2, \ldots, Y_n)\) can be calculated by raising the CDF of a single \(Y_i\) to the power \(n\):
\[ F_{Y_{(n)}}(x) = [F_Y(x)]^n \]
This represents the probability that all of the independent variables \(Y_i\) are less than or equal to \(x\). It's like stacking up these probabilities—only where they all match does \(Y_{(n)}\) achieve this condition which makes this quantity smaller and tighter around the higher values as \(n\) increases.
Remember, the CDF steadily increases from 0 to 1 as \(x\) goes from the smallest to the largest possible values of the variable. Once we differentiate this CDF, we return to find the PDF.

The **Expectation** of a random variable, denoted by \(E\), is essentially the mean or average value that the variable takes on if we were to observe it an infinite number of times. For continuous variables, this expectation is determined through integration. Specifically, for \(Y_{(n)}\) in our example with \(n=2\), we compute it by the formula:
\[ E(Y_{(n)}) = \int_0^1 x \cdot f_{Y_{(n)}}(x) \,dx \]
Expectation helps us by summarizing the overall behavior of the random variable. It gives a single value that can concisely describe typical results from repeated sampling of \(Y_{(n)}\).
Key points are:

• This calculation requires knowledge of the PDF of the distribution.
• The expectation provides a central tendency figure for decision making or predictions.
• It is the "balancing point" of the probability distribution where the outcomes weight each other out.

Thus, computing the expectation over a range takes into account not just the mid-point, but a full weighted average, considering how the distribution spreads across its domain.