A Multianalyzer Machine Learning Model for Marine Heterogeneous Data Schema Mapping
Table 3
Tidal data evaluation factors hierarchy partition.
Objective
Level 1
Level 2
Weights
Evaluation of multisource marine tidal data’s learning machine output
Correlation model evaluation
Conversion coefficient score of the outputs of tide data between and
15%~20%
Matching score of same dimensions tidal data (surface data/underwater data) conversion output
10%~15%
Output distribution probability matching of key data (dynamic location coordinates, time information, wave height information, and tide time information)
5%~10%
Tide height limit spatial matching degree
0%~5%
Conversion performance analysis evaluation
Response time (max, min, and avg) under typical long term observation
5%~10%
Response time (max, min, and avg) under typical short term observation
5%~10%
Response time (max, min, and avg) under long term observation tidal day
0%~5%
Response time (max, min, and avg) under short time observation tidal day
5%~10%
Adaptability analysis evaluation
Data fluctuation under long term observation
5%~10%
Output fluctuation of same dimensions (surface data/underwater data) tidal data
