🐾 PetHealth AI — Pet Disease Classifier
Model Description
Random Forest Classifier that predicts pet health conditions from symptoms. Separately handles Cat and Dog conditions. Trained from scratch on a custom curated veterinary dataset. No pre-trained model or external API used.
Performance
- Accuracy: 94.5%+ on held-out test data
- Train/Test Split: 80/20
- Dataset: 466 custom labeled veterinary cases
- Conditions: 37 across Cat and Dog
Training Details
- Algorithm: Random Forest (100 estimators)
- Encoding: Binary symptom vectors (presence/absence)
- Libraries: scikit-learn, Pandas, NumPy
- Label encoding via LabelEncoder for condition names
How To Load
import pickle
with open("model.pkl", "rb") as f:
model, all_symptoms, pet_type_encoder = pickle.load(f)
with open("label_encoder.pkl", "rb") as f:
le = pickle.load(f)
Links
About
Built by Nabiha Anwar Rana — Rising Sophomore IT Student