Expand description
Machine Learning Pipeline Integration and Real-Time Processing
This module provides a comprehensive ML pipeline framework for SciRS2, enabling real-time data processing, model serving, feature engineering, and automated training workflows.
Features:
- Real-time streaming data processing
- DAG-based pipeline orchestration
- Model serving and inference endpoints
- Feature extraction and transformation pipelines
- Automated model training and evaluation
- Performance monitoring and A/B testing
- Integration with distributed computing and cloud storage
Modules§
- utils
- Convenience functions for common ML pipeline operations
Structs§
- Data
Batch - Batch of data samples for efficient processing
- Data
Sample - Data sample containing features and optional target
- Feature
Schema - Feature metadata and schema information
- Feature
Transformer - Feature transformer for data preprocessing
- MLPipeline
- ML Pipeline orchestrator
- Model
Predictor - Model predictor for inference
- Monitoring
Config - Monitoring configuration
- Pipeline
Config - Pipeline configuration
- Pipeline
Metrics - Pipeline execution metrics
- Streaming
Processor - Real-time streaming processor
Enums§
- Data
Type - Data types supported by the ML pipeline
- Error
Strategy - Error handling strategies
- Feature
Constraint - Feature constraint types
- Feature
Value - Feature value types
- MLPipeline
Error - ML pipeline error types
- Model
Type - Types of ML models supported
- Transform
Type - Types of feature transformations
Traits§
- Pipeline
Node - Pipeline node trait for processing components