Map Matching Evaluator¶
We have implemented several evaluation functions so that different models under the same task can be compared under the same standard.
Evaluation Metrics¶
For the task of map matching, this evaluator implements a series of evaluation indicators:
Evaluation Metrics |
Formula |
|---|---|
RMF(Route Mismatch Fraction) |
\[RMF=(d_{-}+(d_+)/d_0)\]
|
AN(Accuracy in Number) |
\[AN=\frac{\#Rc}{\#R}\]
|
AL(Accuracy in Length) |
\[AL=\frac{\sum len(Rc)}{\sum len(R)}\]
|
where \(d_-\) denotes the length subtracted from the error, \(d_+\) denotes the length added to the error, \(Rc\) denotes correctly matched roads, \(R\) denotes all roads of the real route. \(len(·)\) denotes the length of ·.
Evaluation Settings¶
The following are parameters involved in the evaluator:
Location: libcity/config/evaluator/MapMatchingEvaluator.json
metrics: Array of evaluation metrics,allowed_metricsin evaluator class indicates the type of metrics that the task can accept, andmetricscannot exceed this range.