Machine Learning With Go

Updating pipelines and examining provenance

Over time, we are likely going to want to update our machine learning models. The data that they are processing may have changed over time or maybe we have just developed a different or better model. In any event, it is likely that we are going to need to update our pipeline.

Let's assume that our pipeline is in the state described in the previous section, where we have created the full pipeline, the training process has run once, and we have processed two commits into the attributes repository. Now, let's update the model pipeline to train our multiple regression

Now, let's update the model pipeline to train our multiple regression models instead of our single regression model. This is actually ...