1. Knowledge Hub
  2. Get Started
  3. Train & deploy your own machine learning model

Create an APP

Create a machine-learning application in your project

In this article, you will create a machine-learning application for your project.

An application is a namespace for your feature sets, models, and deployments that is defined by the scope of a specific machine-learning task.

Note: For a more comprehensive definition, see What is an APP? 

1. Create an APP

Within "Your first project", select "Create new APP".

image-Apr-22-2022-07-17-07-17-AM

Step 1: Define the machine learning task

The create app dialog box will open immediately. Here, we want to define our machine-learning application to predict delinquency (classification):

  1. Input application name as: German credit delinquency
  2. Select Data as: german_credit_score - Processed
  3. Select Prediction Task (based on the selected target column, the Engine will automatically assess if it is a regression or classification task)
  4. For the target column, select "SeriousDlqin2Yrs". 
  5. Since in our case, it is a binary classification task, we will also select the positive class label in the target column. In our case, it would just be 1, which indicates that the client in the record will have a serious delinquency event within 2 years of this record.

  6. Click "Next".

define the machine learning class on the AI & Analytics Engine

Step 2: Confirm application configurations

For each machine-learning task, our team of Data Scientists has pre-defined the default configurations. Users are highly recommended to stick to the default configurations unless they have specific and different requirements for the application.

confirm dataset and app configurations

Click "Create Application" and you are done. APP created!

Note: For more details on the APP configurations, see What configurations are in an APP?

 

NEXT: CREATE A FEATURE SET ⇒