The Contribution Score normalizes values and ranks the statistical relationship between dimension items.
The Contribution Score is a normalized value from on scale from -1 to 1 inclusive. Items close to zero are baseline values, indicating that there's no special differences for this item on the date of the anomaly compared to the norm defined in the training period.
Larger values indicate larger differences. A value near 1 indicates the anomaly day is strongly distinct for that reporting item. A value near -1 indicates the anomaly has a complete inverse association.
Complete association for a spike or complete inverse association for a dip.
No association for contribution.
Complete association for a dip or complete inverse association for a spike.
By default, only statistically significant items are returned in Contribution Analysis, derived using Pearson's Residual to calculate a value for each item.