Downloads · 30 days
0
alirezaaminzadeh/unifiedor-domain-classifier
unifiedor-domain-classifier is a text classification model from alirezaaminzadeh. Use it when you need a label for a piece of text. It is set up for sklearn. The card lists the license as apache-2.0.
Lightweight keyword + taxonomy ensemble classifier that predicts the OR domain of a natural-language optimization problem.
Downloads · 30 days
0
Access
Public
Updated Aug 9, 2026
Repo size
—
Likes
0
Public
Click a slice to open those files.
.json2.4 KB · 46%
From the Hugging Face model README
Lightweight keyword + taxonomy ensemble classifier that predicts the OR domain of a natural-language optimization problem.
Given a problem description, predict one of:
Scheduling · Routing · Assignment · Packing · Facility Location · Inventory · Network Flow · Portfolio · Production · Energy
import json
from pathlib import Path
weights = json.loads(Path("classifier_weights.json").read_text())
# Apply keyword + problem_type scoring from weights
Trained on UnifiedOR-100K sample distribution using keyword features, problem-type mapping, and domain priors. Full dataset training runs on HF Jobs alongside dataset build.
| File | Description |
|---|---|
config.json | Model metadata and evaluation metrics |
classifier_weights.json | Trained keyword/prior weights |