BuildingTHEITGUY/ud-campus-parking-occupancy-synthetic

UD Parking Occupancy Classifier (Tiny Demo)

Very small scikit-learn RandomForest classifier trained on the synthetic dataset: BuildingTHEITGUY/ud-campus-parking-occupancy-synthetic

This is a teaching / portfolio model, not a production campus system.

What it predicts

Class label: open | busy | full

Features

capacity, occupied, free, occupancy_ratio, hour_local, weekday

Files

• parking_occupancy_rf.joblib — model artifact
• metrics.json — holdout metrics
• inference_example.py — minimal load/predict script

Quick start

pip install scikit-learn joblib huggingface_hub

from huggingface_hub import hf_hub_download
import joblib

path = hf_hub_download(
    repo_id="BuildingTHEITGUY/ud-parking-occupancy-rf-demo",
    filename="parking_occupancy_rf.joblib",
)
artifact = joblib.load(path)
model = artifact["model"]
classes = artifact["label_encoder_classes"]

# capacity, occupied, free, occupancy_ratio, hour, weekday
x = [[80, 70, 10, 0.875, 10, 1]]
print(classes[model.predict(x)[0]])

Notes / limitations

• Trained only on synthetic rows
• Labels are rule-derived from occupancy_ratio, so reported accuracy is for demo plumbing, not research claim
• No images or personal data included

Author

Mohamed Asath — University of Dubai