Chapter 10: XGBoost Model Deployment
In this final chapter on XGBoost, you will put everything together and develop new techniques to build a robust machine learning model that is industry ready. Deploying models for industry is a little different than building models for research and competitions. In industry, automation is important since new data arrives frequently. More emphasis is placed on procedure, and less emphasis is placed on gaining minute percentage points by tweaking machine learning models.
Specifically, in this chapter, you will gain significant experience with one-hot encoding and sparse matrices. In addition, you will implement and customize scikit-learn transformers to automate a machine learning pipeline to make predictions on data that is mixed with categorical and numerical columns. At the end of this chapter, your machine learning pipeline will be ready for any incoming data.
In this chapter, we cover the following topics:
Encoding mixed data
...
