Statistics & Probability · Level 2 of 5
Entropy
A measure of uncertainty in a probability distribution.
For discrete distributions it is the expected negative log probability of an outcome.
Example
A fair coin has more entropy than a coin that almost always lands heads.
Listen to the definition and example
Audio transcript
Entropy. A measure of uncertainty in a probability distribution. For discrete distributions it is the expected negative log probability of an outcome. For example: A fair coin has more entropy than a coin that almost always lands heads.
Explore this concept
A useful analogy
How much surprise remains before seeing the outcome.
Why it matters
This helps you interpret numerical evidence without overstating what it establishes.
Technical detail
H(p) = −Σ p_i log p_i with 0 log 0 interpreted as zero. Log base two yields bits.
Common misconception
Entropy concerns a distribution, not whether one specific output is correct.
Start with
Related concepts
Quick recall question
Try answering before looking back at the definition.