Reunión 18/5

Introduction to Random Variables

Overview of Random Variables

• The course will cover discrete and continuous random variables, associating discrete variables with counting processes and continuous variables with measuring processes.

• Discrete random variables can take on a finite or countably infinite set of values, which can be related to natural numbers.

Definition and Functionality

• A random variable encodes an experience by assigning numerical values to outcomes in the sample space, forming what is known as the range of the variable.

• For example, when flipping a coin twice, the sample space includes all possible outcomes: heads-heads (HH), heads-tails (HT), tails-heads (TH), and tails-tails (TT). The random variable could represent the number of heads obtained.

Assigning Values to Outcomes

Mapping Sample Space to Real Numbers

• Each outcome in the sample space corresponds to a value assigned by the random variable; for instance, two heads correspond to 2, one head corresponds to 1, and no heads correspond to 0. This forms part of the range of the variable.

Probability Distribution Function

• The probability distribution function assigns probabilities to each value in the range based on how likely it is for that value to occur during experiments. This function takes each element from the range as its domain and outputs a probability value.

Representing Probability Distributions

Methods of Representation

• Probability distributions can be represented through various methods such as:

• Venn diagrams.

• Tables.

• Graphical representations on Cartesian axes.

• Formulas or descriptive language explaining relationships between input and output values.

Key Properties of Random Variables

Conditions for Probabilities

• Two essential conditions must always hold true for any probability distribution:

• Each probability must be greater than or equal to zero.

• The sum of all probabilities across possible outcomes must equal one; if not, there may have been an error in calculations.

Measures Associated with Random Variables

• Important measures include expectation (mean), variance, and standard deviation:

• Expectation represents what one anticipates over many trials.

• Variance indicates how data points differ from their mean.

• Standard deviation is derived from variance but returns units back to original measurements. These measures apply both for discrete and continuous variables.

Calculating Expectation and Variance

Calculation Methods

• Expectation for a discrete random variable is calculated using a summation formula involving each value multiplied by its corresponding probability: E(X) = sum_i x_i P(x_i) .

Variance Formula

• Variance is computed using Var(X) = E(X^2) - (E(X))^2 . It quantifies dispersion around the mean while ensuring results are non-negative due to squaring differences from means before averaging them out.

Properties of Expectation and Variance

Fundamental Properties

• Key properties include:

• The expectation of a constant times a variable equals that constant times the expectation: E(aX) = aE(X) .

• The expectation operator distributes over addition: E(X + Y) = E(X) + E(Y).

• Variances do not share similar properties; they require careful handling especially when dealing with sums or differences involving constants or independent variables.

Example Problem: Expected Value Calculation

Application Scenario

• An example problem involves calculating expected profit based on sales (X) and fixed costs (I). Given their independence, expectations can be combined linearly according to established properties leading towards finding expected profit B defined as B = 20X - I.

Steps Taken

1. Identify given expectations for X & I along with their respective deviations.

2. Apply linearity property in calculating expected profit's expectation using provided formulas without performing explicit calculations during class discussion but emphasizing understanding process instead.

This structured approach allows students not only grasp theoretical concepts but also see practical applications through examples enhancing comprehension significantly throughout learning sessions!

Analysis of Discrete Random Variables

Introduction to the Exercise

• The speaker discusses preparing for a partial defense and emphasizes the need to complete exercises before 8 PM.

• The speaker selects an exercise related to discrete random variables, indicating uncertainty about some calculations but plans to proceed with the task.

Understanding Variance and Standard Deviation

• The discussion includes calculating variance and standard deviation from given probabilities, highlighting that standard deviation is the square root of variance.

Problem Statement on Job Applicants

• A problem involving job applicants is presented: 20% do not know English or computing, 70% meet one requirement, and 10% meet both.

• The variable of interest is defined as the number of requirements met by each applicant.

Constructing Probability Distribution

• The sample space consists of all applicants; the random variable represents how many requirements they fulfill (0, 1, or 2).

• Percentages are converted into whole numbers for constructing a Venn diagram representing applicants' skills in English and computing.

Calculating Probabilities

• The speaker explains how to calculate probabilities based on provided percentages: P(X = 0), P(X = 1), and P(X = 2).

• It’s confirmed that total probabilities must sum up to one. Each probability value is calculated based on previous discussions.

Expectation and Variance Calculation

Steps for Expectation Calculation

• To find expectation (mean), multiply each outcome by its probability: E(X).

Variance Calculation Methodology

• Variance is calculated using E(X² - E(X))². This involves squaring outcomes before multiplying by their respective probabilities.

Transitioning to Binomial Distribution

Introduction to Binomial Distribution Concepts

• A brief overview introduces binomial distribution as a specific type of discrete random variable distribution characterized by independent trials with constant success probability.

Key Parameters in Binomial Distribution

• Two parameters define this distribution: n (number of trials/experiments), and p (probability of success).

Calculating Probabilities in Binomial Distribution

Formula for Probability Calculation

• The formula for calculating binomial probabilities involves combinatorial coefficients multiplied by success and failure probabilities raised to appropriate powers.

Example Problem Setup

• An example problem calculates the probability of finding exactly three defective pieces out of ten produced items with a known defect rate.

Summary Statistics in Binomial Context

• For binomial distributions, expectation (E[X]), variance (Var[X]), and standard deviation can be computed using specific formulas unique to this distribution type.

Practical Application Example

Defining Experiment Parameters

• An example scenario describes taking ten pieces from a production line where there’s a known defect rate. This sets up a classic binomial experiment framework.

Variable Definition Clarification

• Emphasis on correctly defining variables—specifically focusing on counting defective pieces rather than labeling them directly as "defective."

Completing Probability Table

The speaker suggests creating a probability table but acknowledges time constraints while ensuring that all calculated probabilities should sum up to one.

Complementary Probability Approach

• When asked about finding more than two defects, it’s suggested that using complementary probabilities simplifies calculations significantly.

Understanding Probability Distributions

Key Concepts in Probability Calculations

• The speaker discusses the calculation of probabilities, emphasizing that certain values will yield specific results. For example, a value of 10 in one context leads to a result of 1 and remains consistent across different calculations.

• Clarification is made regarding the distinction between "more than two defective pieces" versus "at least two defective pieces." The former starts counting from three, while the latter includes two.

• The speaker explains how to express conditions mathematically, such as defining "two or fewer defects" as X leq 2, which translates into considering values 0, 1, and 2.

• A relationship between events B and C is established; they are complementary events where knowing one helps determine the other. If event B counts more than two defects, event C counts two or fewer.

• The speaker notes that calculating probabilities for both events should sum to one since they cover all possible outcomes within their defined parameters.

Introduction to Poisson Distribution

• The discussion shifts towards the Poisson distribution, identified as a discrete random variable distribution associated with rare events. It’s noted that this is not an exhaustive list of distributions but focuses on those relevant for study purposes.

• Characteristics of the Poisson distribution are outlined: it applies to infinite processes and typically involves calculating occurrences per unit space or time.

• A unique feature of the Poisson distribution is that both its mean (expected value) and variance are equal to lambda (lambda), which represents average occurrences over a specified interval.

Practical Applications of Poisson Distribution

• Examples illustrate practical applications: determining how many service booths should be open at peak times based on expected customer arrivals using Poisson modeling.

• The formula for calculating probabilities in a Poisson distribution is introduced:

[ P(X = k) = fraclambda^k e^-lambdak! ]

where e represents Euler's number (approximately 2.718).

Example Problem Using Poisson Distribution

• An example problem presents itself: determining the probability that at least three customers arrive at a bank in one minute when an average of 120 customers arrive per hour.

• Emphasis is placed on understanding units; since data provided was hourly, adjustments must be made when calculating probabilities for shorter intervals like minutes.

• To find lambda for one minute based on hourly data (120 customers/hour), it must be divided by 60 resulting in lambda = 2.

Calculating Probabilities with Adjusted Lambda

• With adjusted lambda established at 2 for one minute, calculations proceed to determine probabilities for various customer arrival scenarios using cumulative probability methods.

• To find the probability of three or more customers arriving within a minute, it's necessary to calculate P(X geq 3), which can be derived from 1 - P(X < 3).

Final Insights on Expected Values

• When asked about expected customer arrivals over half an hour based on previous calculations (lambda = 120), it’s clarified that this would translate into an expectation of 60, reinforcing that expected values directly correspond with lambda in Poisson distributions.