confidence_code to indicate the confidence level of an AI EDD prediction request. The details are outlined below:
|High confidence for range result, medium confidence for single result
|Medium confidence - Due to force majeure factors (extreme weather, emergencies), etc.
|Medium confidence - Not enough qualified data in this time range
|Medium confidence - shipment_type is not precise
The table above shows the confidence_code that may occur when requesting an estimated delivery date prediction. However, since every customer’s scenario is different we recommend you to choose the method that aligns best with your requirements. However, striking a balance between Accuracy and Coverage is crucial.
- If Accuracy is prioritized, consider accepting only high-confidence results and displaying range EDD. Kindly note that this approach might involve substituting values with your configured EDD or not displaying EDD in certain scenarios.
- If Coverage is more important, opt for all EDD results, even with a medium confidence level result. Enabling EDD dynamic updates can improve the prediction confidence level during the transportation process.
The accuracy of the EDD result is significantly influenced by the input parameters. The confidence result code provides corresponding tips for you:
- If you receive result code 30003, it indicates that you are not specifying your shipment type refer to the shipment type full list to ensure accurate input of the shipment type name for better model understanding.
- If you receive result code 30002, it often signifies that the insufficient training dataset impacts model performance. Providing more related tracking data will assist our AI model in improving. We are excited to hear from you.