Model Card: AutoML Tabular Predictor for Cheese Price

Model Details

• Framework: AutoGluon
• Task: Regression

Dataset

• Source: aslan-ng/cheese-tabular
• Target: price
• Splits:
  • Augmented: 300 rows
  • Original: 30 rows
• Preprocessing Steps:
  • Dropped 'name' and 'origin' columns.
  • Train/test split (80%/20%).

Training

• Framework: AutoGluon
• Preset: "best_quality"
• Time Limit: 300 seconds
• Explored Models: LightGBM, XGBoost, Random Forest, NeuralNetTorch, NeuralNetFastAI, and ExtraTrees.

Best Model

• Model: NeuralNetTorch_r79_BAG_L1
• Time to train: 8.433802 seconds
• Time to inference: 0.109650 seconds
• RMSE Validation: $1.330218
• RMSE Test: $0.869771

Results

• Validation Split:
  • RMSE: $2.0570
  • MAE: $1.5431
  • MSE: $4.2313

Notes