Sign in to save

Bookmark this page so you can find it later.

Sign in to save

Bookmark this page so you can find it later.

Exponential Distribution Reference cheat sheet - grade college

Click image to open full size

Statistics Grade college

Exponential Distribution Reference Cheat Sheet

A printable reference covering the PDF, CDF, mean, variance, memoryless property, and Poisson-process connections for college.

TopicsProbabilityFunctionsGraphsData AnalysisExponentsStatisticsStudy Guide
Download PNG

Related Tools

Related Labs

Related Worksheets

Related Infographics

Study as Flashcards

The exponential distribution models the waiting time until the next event in a process where events occur continuously and independently at a constant average rate. This cheat sheet helps students recognize when the model applies, compute probabilities, and interpret the rate parameter. It is especially useful in reliability, queueing, survival analysis, and Poisson process problems.

The main parameter is the rate λ>0\lambda > 0, which controls how quickly events tend to occur. The density is f(x)=λe−λxf(x)=\lambda e^{-\lambda x} for x≥0x \ge 0, and the cumulative distribution is F(x)=1−e−λxF(x)=1-e^{-\lambda x}. The mean is E[X]=1λE[X]=\frac{1}{\lambda}, the variance is Var⁡(X)=1λ2\operatorname{Var}(X)=\frac{1}{\lambda^2}, and the distribution has the important memoryless property P(X>s+t∣X>s)=P(X>t)P(X>s+t\mid X>s)=P(X>t).

Key Facts

  • If X∼Exp⁡(λ)X\sim \operatorname{Exp}(\lambda), then the probability density function is f(x)=λe−λxf(x)=\lambda e^{-\lambda x} for x≥0x\ge 0 and f(x)=0f(x)=0 for x<0x<0.
  • The cumulative distribution function is F(x)=P(X≤x)=1−e−λxF(x)=P(X\le x)=1-e^{-\lambda x} for x≥0x\ge 0.
  • The survival function is P(X>x)=1−F(x)=e−λxP(X>x)=1-F(x)=e^{-\lambda x} for x≥0x\ge 0.
  • The mean waiting time is E[X]=1λE[X]=\frac{1}{\lambda}, so larger λ\lambda means shorter average waiting time.
  • The variance and standard deviation are Var⁡(X)=1λ2\operatorname{Var}(X)=\frac{1}{\lambda^2} and σ=1λ\sigma=\frac{1}{\lambda}.
  • The memoryless property is P(X>s+t∣X>s)=P(X>t)=e−λtP(X>s+t\mid X>s)=P(X>t)=e^{-\lambda t} for s,t≥0s,t\ge 0.
  • The median is m=ln⁡2λm=\frac{\ln 2}{\lambda} because P(X≤m)=0.5P(X\le m)=0.5.
  • If events follow a Poisson process with rate λ\lambda, then the waiting time until the next event follows Exp⁡(λ)\operatorname{Exp}(\lambda).

Vocabulary

Exponential distribution
A continuous probability distribution used to model waiting time until the next event in a constant-rate process.
Rate parameter
The parameter λ\lambda represents the average number of events per unit time and must satisfy λ>0\lambda>0.
Probability density function
The function f(x)=λe−λxf(x)=\lambda e^{-\lambda x} describes relative likelihood for possible waiting times x≥0x\ge 0.
Cumulative distribution function
The function F(x)=1−e−λxF(x)=1-e^{-\lambda x} gives the probability that the waiting time is at most xx.
Survival function
The function S(x)=P(X>x)=e−λxS(x)=P(X>x)=e^{-\lambda x} gives the probability that the waiting time exceeds xx.
Memoryless property
The rule P(X>s+t∣X>s)=P(X>t)P(X>s+t\mid X>s)=P(X>t) means that the remaining waiting time does not depend on time already waited.

Common Mistakes to Avoid

  • Using the mean as the rate is wrong because λ\lambda is the event rate and E[X]=1λE[X]=\frac{1}{\lambda}, not λ\lambda.
  • Forgetting the support x≥0x\ge 0 is wrong because exponential waiting times cannot be negative, so f(x)=0f(x)=0 for x<0x<0.
  • Confusing P(X>x)P(X>x) with P(X≤x)P(X\le x) is wrong because P(X>x)=e−λxP(X>x)=e^{-\lambda x} while P(X≤x)=1−e−λxP(X\le x)=1-e^{-\lambda x}.
  • Mixing time units is wrong because λ\lambda and xx must use compatible units, such as events per hour with time measured in hours.
  • Assuming every waiting-time problem is exponential is wrong because the model requires independent events occurring at a constant average rate.

Practice Questions

  1. 1 If X∼Exp⁡(0.25)X\sim \operatorname{Exp}(0.25), find P(X>6)P(X>6).
  2. 2 The average lifetime of a component is 500500 hours. Assuming an exponential model, find λ\lambda and P(X≤200)P(X\le 200).
  3. 3 Calls arrive according to a Poisson process at a rate of 33 calls per hour. Find the probability that the next call arrives within 1010 minutes.
  4. 4 Explain why the memoryless property is reasonable for some electronic components but not for a human lifespan.

Understanding Exponential Distribution Reference

A density is not the same thing as a probability at one exact time. Waiting time is continuous, so the chance of an event occurring at precisely one chosen instant is zero. Probabilities come from intervals of time.

The area under the density over an interval gives the chance that the wait falls in that interval. This matters when reading graphs. A high density near zero means short waits are common, but it does not mean a single exact wait has a nonzero chance.

For a time window, it is often easiest to find the chance that no event has happened by the end of the window, then subtract from one. For a window that begins after some time has already passed, compare the probabilities of surviving past each endpoint.

The connection to a Poisson process comes from counting events rather than measuring waits. In a small time interval, the chance of one event is roughly proportional to the interval length. The chance of two or more events in that very small interval is tiny.

Separate intervals do not affect each other. Starting from these ideas, the chance of zero events over a longer interval decreases exponentially with time. That zero-event probability becomes the survival probability for the waiting time.

This link helps students switch between two common problem types. One type asks how many calls, particles, or arrivals occur in a fixed time. The other asks how long until the next call, particle, or arrival.

The memoryless property has a specific meaning. If a bus arrival is truly modeled this way, waiting ten minutes without seeing a bus does not change the distribution of the additional wait. The process has no built-in age.

Its instantaneous event risk stays constant over time. This constant risk is called the hazard rate. Many real systems do not behave this way.

A light bulb may become more likely to fail as it wears out. A patient’s risk can change after treatment. Human arrival rates can rise during rush hour.

In such cases, an exponential model can be a rough approximation but not an exact description. A useful extension occurs when several independent exponential clocks are running.

The first event happens at a rate equal to the sum of their rates. This appears in reliability when any one component can cause system failure.

Units deserve careful attention. If the rate is measured per hour, every time value used in a calculation must be in hours. A rate of two per hour does not mean two hours.

It means an average of two events in one hour, with an average wait of one half hour. When estimating a rate from complete waiting-time data, the usual estimate is one divided by the sample mean wait. Very long waits strongly affect that mean, so a small sample can give an unstable estimate.

Censored data need extra care. For example, if a machine is still working when observation ends, its lifetime is known only to exceed the observation time.

Good model checking compares observed waiting times with the expected exponential shape and looks for changing rates, clusters, or regular spacing. Those patterns suggest that the constant-rate assumption is failing.