20.. Partial Differential Equations

The Normal Distribution is a continuous probability distribution that is symmetric about the mean. It describes data that clusters around a central value with no bias left or right.

The Probability Density Function (PDF) of the normal distribution is given by:

$f(x) = \frac{1}{\sigma \sqrt{2\pi}} e^{-\frac{(x-\mu)^2}{2\sigma^2}}$

Where:

• $x$ = variable
• $\mu$ = mean of the distribution
• $\sigma$ = standard deviation
• $\sigma^2$ = variance

The Normal Distribution is a key concept in statistics. It is defined as a continuous probability distribution that exhibits symmetry about the mean, meaning that the left and right sides of the distribution are mirror images of each other. This characteristic makes it easy to understand how data tends to cluster around a central value, or the mean.

The formula provided is the Probability Density Function (PDF), which gives the likelihood of a variable x taking a specific value. In this function, the mean ($\mu$) determines the center of the distribution, while the standard deviation ($\sigma$) measures the spread of the data. A smaller standard deviation indicates that data points are closer to the mean, while a larger standard deviation signifies more spread or variability in the data.

Imagine the distribution of heights in a large population, like adult males in a city. Most men will cluster around an average height (the mean), say 70 inches. Very few will be extremely short or extremely tall. This creates a bell-shaped curve if we were to graph their heights, where the majority are near the average and fewer at the extremes. This is essentially what the Normal Distribution resembles.

1. Bell-shaped curve: Symmetrical about the mean.
2. Mean = Median = Mode: All measures of central tendency are equal.
3. Total area under the curve = 1: This represents the total probability.
4. Empirical Rule (68-95-99.7 Rule):
   • 68.27% of data lies within $\mu \pm \sigma$
   • 95.45% within $\mu \pm 2\sigma$
   • 99.73% within $\mu \pm 3\sigma$
5. The curve never touches the x-axis: It extends infinitely in both directions.

The properties of the Normal Distribution help to summarize the essential features of the data:

1. Bell-shaped curve: This indicates that as values move away from the mean, the probabilities of those values occurring decrease.

2. Central tendencies: The Normal Distribution is unique because the mean, median, and mode are all the same value, highlighting that the data is symmetrically distributed.

3. Total area under the curve: A fundamental property of probability distributions is that the total area under the curve equals 1. This assures that all possible outcomes together represent the entire probability space.

4. Empirical Rule: This rule simplifies calculating probabilities within a normal distribution. It states:

   • About 68% of the data falls within one standard deviation from the mean,
   • About 95% falls within two standard deviations,
   • About 99.7% falls within three standard deviations. This is crucial in assessing the variability of a dataset.

5. Infinitely extending curve: The fact that the curve never touches the x-axis indicates that there is always a non-zero probability for extreme values, ideally making it applicable across a wide range of data.

Think of exam scores for a large class: if most students score around the average with few scoring much higher or lower, the distribution of scores resembles a bell curve. For example, if the average score is 75, and the standard deviation is 10, then about 68% of students will score between 65 and 85, while almost all will score between 55 and 95. This helps teachers understand overall student performance and set grading scales.

The Standard Normal Distribution is a special case of the normal distribution with:

• Mean $\mu = 0$
• Standard deviation $\sigma = 1$

To convert a normal distribution to a standard normal distribution, we use the Z-score formula:

$Z = \frac{X - \mu}{\sigma}$

Where:

• $X$ = any value from the original normal distribution
• $Z$ = corresponding value in the standard normal distribution

The Standard Normal Distribution simplifies calculations in statistics. It is a specific form of the Normal Distribution where the mean is set to 0 and the standard deviation to 1. This allows for easier comparisons between different normal distributions.

When we need to convert a value from any normal distribution into the standard form, we employ the Z-score formula. The Z-score tells us how many standard deviations a given value is from the mean. This is helpful in identifying the relative standing of that value within its dataset.

Consider measuring the temperature in various cities around a country. If you want to compare a specific city's temperature to the national average using temperatures from multiple cities, converting these temperatures into Z-scores allows you to see which cities are colder or hotter in relation to the average, regardless of the original scale of measurement.

• Engineering: Noise distribution in electronics, error modeling in control systems • Manufacturing: Quality control and Six Sigma process control • Finance: Stock return distributions, risk analysis • Biology/Medicine: Heights, weights, test scores • Machine Learning: Assumptions in probabilistic models

The Normal Distribution is instrumental across various fields due to its properties:

• Engineering: In electronics, engineers often analyze the noise produced by circuits, which can be modeled as normally distributed. Understanding this can help mitigate errors in control systems.
• Manufacturing: The quality control processes utilize the principles of normal distributions, particularly in Six Sigma, which strives to reduce defects by understanding variability.
• Finance: Analysts use the normal distribution to model stock returns. They evaluate the risk of investments and future returns by understanding where their expected returns lie in relation to historical data.
• Biology/Medicine: Normal distributions often characterize biological measurements such as height, weight, or scores on medical tests, helping in determining norms and outliers in health assessments.
• Machine Learning: Many algorithms rely on the normality assumption to predict outcomes, making this distribution crucial for creating more accurate models.

Think of the manufacturing process in making footballs. If a company desires that their footballs weigh 500 grams on average, they must account for the normal distribution of potential weights. By ensuring that most footballs fall within a specified range of weight (e.g., between 490 and 510 grams), they can maintain quality and performance standards in their product line, therefore satisfying customers.

Step-by-Step Procedure

1. Identify: Mean ($\mu$) and Standard deviation ($\sigma$)
2. Convert: Convert raw score to Z-score using the formula
3. Refer: Use Z-table (standard normal table) to find area (probability)
4. Interpret: Convert area into required probability

When faced with a problem involving the Normal Distribution, you can follow a structured approach:

1. Identify the mean and standard deviation of your dataset. These parameters define the distribution.
2. Convert raw scores to Z-scores using the formula, which standardizes the score for comparison.
3. Refer to the Z-table to find the area corresponding to your Z-score. This area represents the probability of observing a value less than that one.
4. Interpret this area in the context of your problem, which may involve calculating probabilities or assessing outcomes based on certain thresholds.

Let's say a teacher has the scores of a math test where the average score was 75 with a standard deviation of 10. If a student scored 85, the teacher wants to find out the likelihood that a student scores less than 85:

1. Identify: $\mu = 75$, $\sigma = 10$.
2. Convert: The Z-score would be $Z = \frac{85 - 75}{10} = 1$.
3. Refer: From the Z-table, we find that $P(Z < 1) = 0.8413$.
4. Interpret: This means about 84.13% of students scored below 85.

Example 1:

The marks of students in a class are normally distributed with a mean of 70 and standard deviation of 10. What is the probability that a student scores less than 85?

Solution:

• Given: $\mu = 70$, $\sigma = 10$, $X = 85$
• Convert to Z: $Z = \frac{85 - 70}{10} = 1.5$
• From Z-table: $P(Z < 1.5) = 0.9332$

Answer: Probability = 0.9332 or 93.32%

Example 2:

Find the probability that a value lies between 60 and 80.

Solution:

• $\mu = 70$, $\sigma = 10$
• Convert both values to Z: $Z_{60} = \frac{60 - 70}{10} = -1$ $Z_{80} = \frac{80 - 70}{10} = 1$

From Z-table:

• $P(Z < 1) = 0.8413$
• $P(Z < -1) = 0.1587$

$P(60 < X < 80)= 0.8413 - 0.1587 = 0.6826$

Answer: 68.26%

The examples demonstrate practical applications of the Normal Distribution:

• Example 1 walks through finding the probability of obtaining a score lower than 85, given the class's mean and standard deviation. By identifying the parameters, converting the value to a Z-score, and referencing the Z-table, we find a substantial likelihood of 93.32%.

• Example 2 explores finding the probability of scores falling within a range, by converting the values at both ends of the interval to Z-scores, checking the Z-table for the respective probabilities, then calculating the difference to find the probability of the range itself.

Consider the scenario of predicting customer wait times in a busy restaurant. If you gathered wait times and found that they followed a normal distribution with an average wait time of 30 minutes, and a standard deviation of 5 minutes.

• If you’d like to know the likelihood a customer waits less than 35 minutes, you could solve it using a similar approach as in Example 1. Or, if you want to know the probability of customers waiting between 25 and 35 minutes, you can apply the methodology shown in Example 2. This not only satisfies customer service insights but also allows the business to manage staff and expectations better.

For large values of $n$ (usually $n > 30$), the binomial distribution can be approximated using the normal distribution: $X \sim B(n,p) \Rightarrow N(\mu = np, \sigma = \sqrt{npq})$

Where:

• $p$ = probability of success
• $q$ = 1 - $p$

Apply continuity correction by adjusting values by ±0.5 when switching from discrete to continuous.

The Normal Approximation to the Binomial Distribution allows statisticians to apply the properties of the normal distribution to binomial data, especially when the number of trials (n) is large (greater than 30). This is advantageous because calculating probabilities with the normal distribution is typically simpler than with the binomial formula.

The transition from the binomial to the normal distribution involves finding the mean (np) and the standard deviation ($\sqrt{npq}$), where q represents the probability of failure.

However, it’s important to apply a continuity correction by adding or subtracting 0.5 when estimating probabilities over a range; this accounts for the fact that we are transitioning from discrete to continuous modeling.

Suppose a factory produces light bulbs, and we know the probability of a bulb being defective is 0.1. By observing a large batch of 1000 bulbs, we can utilize normal approximation rather than doing numerous binomial calculations. If we want to find the chance of discovering no more than 100 defective bulbs, we can convert our problem into the normal distribution framework without dealing with complex binomial calculations, making it more manageable.

• It assumes symmetry; real-world data may be skewed. • It’s sensitive to outliers. • Not suitable for data bounded on one side (e.g., wait times, length of objects).

While the Normal Distribution is powerful, it has limitations:

• Assuming symmetry: Real-world data might not always be symmetric. For example, income distribution often skews right with a long tail on higher incomes.
• Sensitivity to outliers: Extreme values can heavily influence the mean and standard deviation, skewing the distribution and its reliability.
• Not suitable for one-sided data: For datasets that have a natural bound on one side, such as waiting times or lengths (which can’t be negative), the normal distribution may lead to inaccurate conclusions.

Consider the time it takes for customers to be served in a coffee shop. If some days are exceptionally busy, leading to a few outliers who wait much longer than the typical customers, the average wait time might be skewed higher than most expect. Instead of assuming a normal distribution, the owner might need to acknowledge these extremes to set realistic wait time expectations and improve service efficiency.

--