This article explains the concept of build from scratch in the AI & Analytics Engine.
The AI & Analytics Engine is used to solve business problems in various ways. One of them is by using apps.
If you want to learn more about apps, see this article.
There are 2 ways to create an app on the platform:
-
Build from scratch.
This article will describe the “build from scratch” option.
What is “build from scratch”?
“Build from scratch” is the “manual” method (compared to “build from template”) for creating either a regression, classification or a clustering app.
For more information about these types of apps, see the articles below:
1) What is classification?
2) What is regression?
3) What is clustering?
Unlike templates, it is not related to any specific business problem. Therefore, this option is able to support a larger variety of problems.
It also gives the user more flexibility over various elements in the process of creating a prediction pipeline. For example, allowing control over data preparation recipes or enabling manual selection of models for training, etc.