case2/prep/02Unsupervised_Machine_Lear.../q

11 lines
332 B
Plaintext

Normal machine learning: Input - Label
Here: Only Input.
->Instead of classifying different types, try to understand your given dataset
Deviations from this understanding are anomalies
<l2st>
x: training samples
tx: test samples
ty: test labels (is a certain sample an anomaly or not)
</l2st>
Useful: \emph{peak /global/cardio.npz}