βοΈ Indian Domestic Flight Price & 'Buy vs. Wait' Advisor Model
Trained on 300,153 Indian domestic flight records covering top metro routes (DEL, BOM, BLR, CCU, HYD, MAA) to predict fair market fares in INR (βΉ) and deliver real-time booking advisory.
π Model Performance
- Algorithm: LightGBM Regressor (Gradient Boosted Decision Trees)
- $R^2$ Score:
0.9827(98.3% variance explained) - MAPE:
12.89% - MAE:
βΉ2,717.21 - RMSE:
βΉ4,933.84
π» How to Use in Python
from huggingface_hub import hf_hub_download
import joblib
import pandas as pd
# 1. Download Model from Hugging Face Hub
model_path = hf_hub_download(repo_id="ranjeet447/indian-flight-price-advisor", filename="model.joblib")
model = joblib.load(model_path)
# 2. Download Feast Route Stats
features_path = hf_hub_download(repo_id="ranjeet447/indian-flight-price-advisor", filename="route_features.parquet")
route_stats = pd.read_parquet(features_path)
print("β
Model and Feast features loaded successfully!")
Inference Providers NEW
This model isn't deployed by any Inference Provider. π Ask for provider support