Intro
ecosystem.Ai uses a worker architecture to define and run predictions. The base configurations are stored in the ecosystem_meta
database across many collections.
Key Features
- Meta-Data Driven: All predictor configuration items are managed from the model meta-data.
- Broad Range of Parameters: Parameters are model dependent and can be set in the model meta-data.
- Many Model Types: Multiple model types are supported, including regression, classification, and clustering.
- Model Result: The model result is stored in the
models
database. - Easy Testing of Models: Models can be tested using the
Test Model
option. - Deploy: Models are saved in the
models
database and can be deployed for predictions.