Machine Learning Workflow

We shall automate the whole process as much as possible. Pack these steps in one pipeline and run the pipeline with one command.

In fact it’s an iterative process: once we get the pipeline up and running, we can anaylze outputs of some steps to gain more insight, and then come back to the whole exploration steps to make some improvement. Repeat the process until we get an accepted performance.

Data Collection and Visualization,
Data Preparation, Data Set Split, Pipeline,
Cross Validation, Fine-tune the model,
Performance Measures, Improvement,