Skip to main content

XGBoost Regression

Gradient boosted decision trees for regression tasks on tabular data.

  • Model ID: xgboost-regression (CPU). A GPU histogram-training variant, xgboost-regression-gpu, is also available for large datasets.
  • Multi-output: Yes (predicts multiple target columns simultaneously)
  • Hardware: CPU (default); GPU via the -gpu variant
  • Specialist type: tabular

For classification use xgboost-classification. The full list of classical model IDs is in the ML Classical index.

When to Use​

  • House price prediction
  • Sales forecasting
  • Demand prediction
  • Revenue estimation
  • Multi-output regression (predict multiple targets at once)

Quick Start​

Single-output​

from colabhive import ColabHive

client = ColabHive(api_key="your_api_key")

dataset = client.datasets.upload(
    name="my_dataset",
    file="./data.csv"
)

job = client.training.create(
    model="xgboost-regression",
    dataset_id=dataset.id,
    hyperparameters={
        "target_column": "price"
    }
)

job.wait()
print("Training completed!")
print(job.get_metrics())   # {"mae": ..., "rmse": ..., "r2": ...}

Multi-output​

job = client.training.create(
    model="xgboost-regression",
    dataset_id=dataset.id,
    hyperparameters={
        "target_columns": ["price", "demand", "rating"]
    }
)
Multi-output

When using target_columns (plural), XGBoost uses MultiOutputRegressor to train one model per target efficiently. Metrics are reported per target and aggregated.

Try with Example Dataset​

Dataset Requirements​

Supports:

  • CSV (.csv)
  • JSONL (.jsonl)
  • Parquet (.parquet)

Your dataset should be tabular:

  • Features: numeric or categorical columns
  • Target: one column (single-output) or multiple columns (multi-output)
  • Minimum samples: 50 (recommended: 1000+)

Example:

feature1,feature2,feature3,target
1.2,3.4,5.6,10.5
2.3,4.5,6.7,15.2

Expected Results​

Regression metrics:

  • MAE (lower is better)
  • RMSE (lower is better)
  • R² (higher is better; 1.0 is perfect)

Technical Details​

  • Framework: xgboost
  • Category: ml_classical
  • Hardware: CPU (xgboost-regression) or GPU (xgboost-regression-gpu)