haipipe-nn-1-algorithm

What problem does it solve?

This Skill addresses the challenge of integrating diverse external Machine Learning algorithms into a unified pipeline, abstracting away their individual complexities.

Core Features & Use Cases

• Algorithm Abstraction: Wraps external ML libraries (e.g., XGBoost, NeuralForecast, Transformers) into a consistent Tuner interface.
• Diversity Handling: Manages variations in input/output formats, training paradigms, and serialization methods across different algorithms.
• Use Case: A data scientist can use this Skill to easily swap between different forecasting algorithms (like PatchTST or ARIMA) without altering the core pipeline logic, simply by configuring the appropriate Tuner.

Quick Start

Configure the pipeline to use the XGBoost algorithm for model training.