121
edits
Changes
File:Ml pipeline fr.png (view source)
Revision as of 17:06, 10 January 2023
, 17:06, 10 January 2023no edit summary
Depiction of a three-step process for the machine learning pipeline. The first step is data preparation. This is described as transforming the CTD data into features for a machine learning model. The second step is data partitioning. This is described as using a Gaussian Mixture Model to partition data into areas of differing water properties. The third step is data classification. This is described as using a Multi-Layer Perceptron to flag CTD scans for deletion or preservation.