Event
An event is just “something you care about happening” when there’s uncertainty. If you repeat a random process—like drawing a card, measuring a patient’s blood pressure, or seeing whether a user clicks an ad—an event is the particular outcome (or set of outcomes) you want to talk about.
What an event is, technically
In probability, you start with a sample space: the set of all possible outcomes. An event is any subset of that sample space. That means events can be simple (“the die shows 6”) or broad (“the die shows an even number”). The probability of an event, written P(A), is a number between 0 and 1 measuring how likely it is.
Everyday examples
- House prices: A = “a listed home sells above $500k.”
- Exam scores: B = “a student scores at least 90.”
- Medical tests: C = “the test result is positive.”
- Online ads: D = “the user clicks within 10 seconds.”
Each of these events groups outcomes into “yes, it happened” vs “no, it didn’t,” which is exactly what probability needs.
How events combine
Events follow set logic:
- Union (A ∪ B): A happens or B happens.
- Intersection (A ∩ B): A happens and B happens.
- Complement (Aᶜ): A does not happen.
These operations power rules like conditional probability P(A|B) and ideas like independence.
Why it matters in AI/ML
Many ML tasks are “event prediction”: P(y=1|x) is the probability that an event (spam, fraud, disease) occurs given features. Loss functions like log loss, models like logistic regression, and metrics like ROC-AUC all revolve around assigning good probabilities to events.
Event is a set of outcomes of a random experiment, typically represented as a subset of the sample space. Events are the basic objects to which probabilities are assigned, enabling formal reasoning about uncertainty in statistics and AI/ML. They support defining likelihoods, losses, and decision rules. Example: in binary classification, the event “model predicts positive” can be analyzed to estimate error rates such as false positives.
Imagine you’re flipping a coin. You don’t know what will happen, but you can talk about certain “things that might happen,” like “it lands on heads” or “it lands on tails.” Each of those is an event.
In probability (and in AI/ML), an event is simply a specific outcome or a set of outcomes you care about from some uncertain process. For example, in a spam filter, an event could be “this email is spam,” or “the model predicts spam.” Probabilities are assigned to events to describe how likely they are.