Biểu đồ ven và xác suất
Xác suất có điều kiện và xác suất tích
When using experimental probabilities with large datasets, the likelihood of the events being statistically independent is negligible. This means that any two events can easily be shown to be dependent, for example, "a two is rolled on a dice" and "the telephone will ring in the next three minutes".
To avoid the conclusion that unrelated events are dependent, we can say events are independent when P(A|B) is within x percent of P(AnB)/P(B), instead of requiring them to be exactly equal.
Drag the blue slider to increase x, i.e. to make statistical independence more achievable.