Building and tuning your model using JupyterHub

As a data scientist, you will want to try different models with different parameters to find the right fit. Before you start building the model, recall from Chapter 8, Building a Complete ML Project Using the Platform, that you need to define the evaluation criteria, and that accuracy may be a misleading criterion for a lot of use cases.

For the flight use case, let's assume that your team and the SME agree on the PRECISION metric. Note that precision measures the portion of correct positive identification in the provided dataset.

Let's start writing our model and see how the platform enables data scientists to perform their work efficiently:

1. Open the chapter10/experiments.ipynb file notebook in your JupyterHub environment.
2. In Cell 2, add the connection information to MLflow. Recall that MLflow is the component in the platform that records the model experiments and works as the model registry. In the code,...