Continuous Probability Distributions
So far, we studied discrete probability distributions, where values could be counted.
In this lesson, we move to continuous probability distributions, used when values are measured, not counted.
Continuous distributions are essential in statistics, physics, economics, data science, machine learning, and real-world measurements.
Why Continuous Distributions Are Needed
Many real-world quantities are not whole numbers.
Examples include:
- Height of a person
- Time taken to complete a task
- Weight of a package
- Temperature
These values can take infinitely many possibilities. Discrete models cannot handle them properly.
What Is a Continuous Random Variable?
A continuous random variable can take any value within a given interval.
Between any two values, there are infinitely many possible values.
Because of this, we do not assign probabilities to exact values.
Key Idea (Very Important)
For a continuous random variable X:
P(X = a) = 0
Probability is assigned to intervals, not exact points.
This concept is critical for exams and understanding.
Probability Density Function (PDF)
A Probability Density Function (PDF) describes how dense the probability is around a value.
It is denoted by:
f(x)
The PDF itself is NOT a probability.
How Probability Is Calculated (Visualization Idea)
For continuous distributions:
Probability = Area under the curve
between two values.
This is why graphs are extremely important here.
Properties of a PDF
Every valid probability density function must satisfy:
- f(x) ≥ 0 for all x
- Total area under the curve = 1
If these conditions fail, it is NOT a valid PDF.
Example: Uniform Continuous Distribution (Idea)
In a continuous uniform distribution:
- All values in an interval are equally likely
- The graph is a flat horizontal line
Example:
Waiting time between 0 and 10 minutes.
Probability depends on interval length.
Probability Over an Interval (Concept)
If X is continuous, probability that X lies between a and b is:
P(a ≤ X ≤ b)
This equals the area under the PDF from a to b.
This replaces counting logic.
Discrete vs Continuous (Key Comparison)
| Aspect | Discrete | Continuous |
|---|---|---|
| Values | Countable | Uncountable |
| Probability at a point | Possible | Zero |
| Function | PMF | |
| Visualization | Bars | Curve |
This table is frequently tested in exams.
Cumulative Distribution Function (CDF)
The Cumulative Distribution Function (CDF) gives the probability that:
X ≤ x
It is defined as:
F(x) = P(X ≤ x)
CDF always increases from 0 to 1.
Graphical Meaning of CDF
The CDF at a value x represents the total area under the PDF curve to the left of x.
This gives cumulative probability.
Examples of Continuous Distributions
- Uniform distribution
- Normal distribution
- Exponential distribution
Each models different real-world behavior.
Continuous Distributions in Daily Life
Examples include:
- Time taken to travel to work
- Heights of students
- Temperature during a day
Exact values are unpredictable, but ranges are meaningful.
Continuous Distributions in Competitive Exams
Exams test:
- PDF properties
- Difference between PMF and PDF
- Interval-based probability logic
Remember: point probability is zero.
Continuous Distributions in Statistics
Statistics uses continuous distributions to:
- Model natural phenomena
- Compute probabilities using areas
- Analyze variability
Most real measurements are continuous.
Continuous Distributions in Data Science
In data science:
- Features like time, price, age
- Error terms
are modeled using continuous distributions.
Continuous Distributions in Machine Learning
Machine learning uses continuous distributions in:
- Regression models
- Loss functions
- Probabilistic models
Gaussian assumptions are very common.
Common Mistakes to Avoid
Students often:
- Try to find P(X = a)
- Confuse PDF with probability
- Forget area interpretation
Always think in intervals and areas.
Practice Questions
Q1. Can P(X = 5) be non-zero for a continuous variable?
Q2. What represents probability in continuous distributions?
Q3. What must the total area under a PDF equal?
Quick Quiz
Q1. Is PDF itself a probability?
Q2. Does continuous data use curves for visualization?
Quick Recap
- Continuous variables take uncountable values
- Probability at a point is zero
- PDF describes density, not probability
- Probability = area under the curve
- Foundation for normal distribution
With continuous distributions understood, you are now ready to explore the Expected Value and Variance, which quantify average behavior and spread.