skills.rest
olavocarvalho
olavocarvalhoCOMMUNITY
Explore all skills by this author →
𝕏@olavovieiracomGitHub

tabular-ml-modeling

Develops GPU-accelerated gradient boosting models for large-scale tabular data and classification tasks.

1|Updated Jan 26, 2026
Search Tags:#machine learning#GPU acceleration#cross-validation#tabular data#gradient boosting#ensembling
One-click install
npx skills add https://github.com/olavocarvalho/data-agents --skill tabular-ml-modeling
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: tabular-ml-modeling
Source: https://github.com/olavocarvalho/data-agents/tree/main/skills/tabular-ml-modeling
Command: npx skills add https://github.com/olavocarvalho/data-agents --skill tabular-ml-modeling

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires cudf-cu12, cuml-cu12, xgboost, catboost, lightgbm, and includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve?

This Skill tackles the challenge of building high-performance machine learning models for large-scale tabular datasets, enabling accurate predictions for regression and classification tasks.

Core Features & Use Cases

  • GPU-Accelerated Training: Leverages XGBoost, LightGBM, and CatBoost on NVIDIA A100 GPUs for rapid iteration.
  • Advanced Validation: Implements era-based cross-validation to prevent temporal leakage, crucial for time-series data.
  • Ensemble Techniques: Combines diverse models using methods like hill climbing and stacking for improved robustness and accuracy.
  • Use Case: Predict stock market movements or customer churn by training sophisticated models on millions of rows of structured data, optimizing for competition-winning metrics.

Quick Start

Use the tabular-ml-modeling skill to train an XGBoost model on the provided training data, validating using era-based cross-validation.

Frequently Asked Questions about tabular-ml-modeling

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I train machine learning models on large-scale tabular data?▼

To train machine learning models on large-scale tabular data, you can use GPU-accelerated gradient boosting frameworks like XGBoost, LightGBM, and CatBoost. This approach handles millions of rows and thousands of features efficiently for regression and classification tasks.

Can I use XGBoost and LightGBM with GPU acceleration for structured datasets?▼

Yes, you can use XGBoost and LightGBM with GPU acceleration for structured datasets. The setup leverages RAPIDS acceleration libraries like cuDF and cuML, optimized specifically for NVIDIA A100 GPUs with CUDA 12.4.

What is era-based cross-validation in gradient boosting?▼

Era-based cross-validation is an advanced validation strategy that prevents temporal leakage in time-series data. It ensures that your gradient boosting models are evaluated on distinct chronological eras, which is crucial for accurately predicting stock market movements.

How do I combine multiple gradient boosting models using ensembling?▼

You combine multiple gradient boosting models using ensembling techniques like hill climbing and stacking. This method merges diverse models trained on tabular data to improve overall robustness and prediction accuracy.

Do I need an NVIDIA A100 GPU to run tabular machine learning workflows?▼

You need an NVIDIA A100 GPU with CUDA 12.4 to utilize the RAPIDS acceleration and GPU-optimized gradient boosting frameworks as intended. This hardware configuration enables the system to handle millions of rows and thousands of features efficiently.

What's the best way to predict stock market movements using tabular data?▼

The best way to predict stock market movements using tabular data is training sophisticated gradient boosting models with era-based cross-validation. This prevents temporal leakage and optimizes competition-winning metrics on millions of structured rows.

Related Skills

Explore More Data & Analytics Skills→
coinvest518coinvest518

cudf-analytics

Analyze large tabular datasets with NVIDIA cuDF for GPU-accelerated grouping, summarization, and anomaly detection.

Community
Intermediate
gizixgizix
2

model-deployment

Package trained machine learning models with versioning, ONNX export, and FastAPI endpoints.

Community
Advanced
bogwarebogware
3

cuml-machine-learning

Train and evaluate cuML classification, regression, clustering, and dimensionality reduction models on tabular datasets.

Official
Intermediate
leesk212leesk212
1

cuml-machine-learning

Train and evaluate GPU-accelerated machine learning models on tabular data with cuML.

Community
Intermediate