In this lab, you will use SparkML in Watson Studio to run generated travel data through a machine learning algorithm, automatically tune the algorithm, and save the model into the project repository.
Upon completing the lab, you will know how to:
- Create a projec token.
- Identify labels and transform data.
- Conduct feature engineering for algorithm data.
- Declare a machine learning model.
- Setup the Pipeline for data transforms and training.
- Train the data.
- Show and evaluate machine learning results.
- Automatically tune machine learning results.
- Save the model to the model repository