Tutorials Guide
๐ Sign Up for Our Hosted Tutorials
Experience hands-on learning with our Hosted Tutorials. Gain access to our hosted environment and explore FeatureByte in an interactive setting.
Learn to Use the FeatureByte User Interface¶
What Youโll Learn¶
-
Catalog Creation: Organize and manage your data sources in a structured catalog.
-
Data Modeling: Define tables and entities to support efficient feature computation.
-
Use Case Structuring: Design and manage feature engineering workflows around business use cases.
-
Ideation: Discover how FeatureByte automatically suggests relevant features and models based on your data and objectives.
-
Experimentation: Build and iterate on feature lists, view models' key metrics, and select the best model for deployment.
-
Deployment: Operationalize your final feature list and model seamlessly.
๐ Dataset Overview¶
The Credit Default Dataset includes seven tables:
| Table Name | Description |
|---|---|
| NEW_APPLICATION | Information about new loan applications. |
| CLIENT_PROFILE | Client demographic and profile information. |
| BUREAU | Records of previous credits taken by clients from other institutions. |
| PREVIOUS_APPLICATION | Details of prior loan applications by each client. |
| INSTALLMENTS_PAYMENTS | Logs of monthly installment payments for loans. |
| LOAN_STATUS | Current loan status over time. |
| CREDIT_CARD_MONTHLY_BALANCE | Monthly balance summaries for clientsโ previous credit cards. |
๐น Start Here¶
Learn to Use the FeatureByte SDK¶
Step 1: Set Up the SDK¶
Follow the SDK Tutorial Installation Guide for step-by-step setup instructions.
Step 2: Build an End-to-End Workflow¶
Using the same dataset as in the UI tutorials, learn to:
- Create a catalog and register your data model
- Define use cases and craft features
- Compute training data and train models
- Deploy and manage production-ready features
๐น Start Here¶
Step 3: Work with Item Data in the SDK¶
๐ Dataset Overview¶
The Grocery Dataset includes four tables:
| Table Name | Description |
|---|---|
| GroceryCustomer | Customer details such as name, address, and date of birth. |
| GroceryInvoice | Invoice details including timestamps and total amounts. |
| InvoiceItems | Items in each invoice, with quantities, discounts, and costs. |
| GroceryProduct | Product group and description for grocery items. |
๐น Start Here¶
Tutorial Differences
Both datasets include Event Tables, Slowly Changing Dimension Tables, and Dimension Tables.
| Dataset | Additional Table Type |
|---|---|
| Grocery Dataset | Item Table |
| Credit Default Dataset | Time Series Table |
Learn to Implement Custom Transformers¶
Enhance your text processing and feature engineering capabilities by integrating custom transformer models. Explore our Bring Your Own Transformer guide to learn how.
Learn to Export Features as SQL Code¶
Export FeatureByte features as standalone SQL templates for integration into custom pipelines. Use FeatureByte Feature Ideation to discover features, then export and run them independently in your data warehouse.