Knowledge Maxxing← Back to topic library
Responsible Tech MaxxingIntermediateBias75 sec
Dataset bias
Dataset bias is a systematic imbalance or gap in a training dataset that causes a model to perform unevenly across different groups or situations.
What matters
Because a model can only learn from the data it's given, a biased dataset directly limits how fairly or accurately the resulting model can treat underrepresented groups.
Example
A facial recognition dataset containing mostly light-skinned faces can produce a model that performs noticeably worse on darker-skinned faces.
Common misconception
Dataset bias is not fixed simply by collecting more data of the same kind; more data with the same imbalance just reinforces the existing gap.
Recall question
Why doesn't collecting more data automatically fix dataset bias?
Keywords and related concepts
dataset biasunderrepresentationtraining datasampling bias