Plan

First, you have to define a plan of what you want to achive with your data science project. To do this, we start with the business model, which describes the rationale of how your organization creates, delivers and captures value. The complete process includes the following topics:

  1. Identify use case: Use the business model canvas.

  2. Frame the problem: Provide a statement of what is to be learned and how decisions should be made.

  3. Identify variables or labels: for structured data problems, we need to identify potentially relevant variables; for unstructured data problems, we need to define labels.

  4. Define success metrics: Write down your metrics for success and failure with the data science project.

To learn more about the data science planning phase, review the following presentations.

Identify use case

Use the business model canvas.


Frame the problem

Provide a statement of what is to be learned and how decisions should be made.


Identify variables

For structured data problems, we need to identify potentially relevant variables; for unstructured data problems, we need to define labels.


Define success metrics

Write down your metrics for success and failure with the data science project.