Decoder

Dataset: a table of information, one row per thing you measured.

Average (mean): add everything up and divide by how many there are. One huge value can drag it far from the crowd.

Median: line everything up and take the middle one. Much harder to push around.

Outlier: a value far away from the rest. It might be an error, or the most interesting row in the file.

Correlation: how closely two things move together, from -1 to +1. Moving together does not mean one causes the other.

Bias: a pattern in how data was collected that makes it unfair or misleading.

Training: showing a machine many examples so it can learn a pattern.