Research Article
Online Bridge Structural Condition Assessment Based on the Gaussian Process: A Representative Data Selection and Performance Warning Strategy
Algorithm 1
General framework for representative data selection.
| (1) | Input: new data point | | (2) | Persistent: current subset selection , hyperparameters of the GP model | | (3) | Compute the novelty of data point | | (4) | If , then | | | Keep the current subset and GP model | | (5) | Else | | | Adding to and update the GP model | | (6) | End if | | (7) | If data size exceeds , then | | | Compute the score (i.e., redundancy, ) of each datapoint based on its impact of removal | | (8) | Remove the data point with the lowest score (i.e., minimum impact of removal, ) from the dataset and update the GP model | | (9) | End if |
|
