1{"record_id": "41c05e4c190bfd812f576bb4bc53a0c147331a7d", "eligible": true, "short_ref": "Deppner & Cajias (2024)", "country": "Germany", "city": "Frankfurt am Main", "property_type": "apartments (rental listings)", "price_type": "asking rent (monthly, log-transformed)", "n_obs": 9256,2 "sample_period": "2019-01 to 2020-03", "data_source": "German multiple listing systems (MLS) apartment rental listings", "methods_compared": ["OLS hedonic", "SDM (spatial Durbin)", "SDEM (spatial Durbin error)", "random forest", "XGBoost"],3 "best_method": "XGBoost", "metric": "R2", "best_value": 0.8521, "hedonic_value": 0.7881,4 "test_set": "holdout", "features_ai": "none",5 "incremental_ai_gain": null, "incremental_ai_metric": "", "marginal_effects": "coefficients", "notes": "Table 4 Panel A (models with spatial controls): true out-of-sample holdout = Q1 2020 listings predicted from models trained on 2019 data. R2 printed as 85.21% for XGB (spatial tuning; 85.20% non-spatial tuning), 85.13% RFR, 81.62% SDM, 78.81% OLS; MAPE 12.83% XGB vs 15.03% OLS; RMSE 244.06 vs 292.14 EUR/month. Paper's main point is that non-spatial k-fold CV errors of RF/XGB are over-optimistic (CV R2 ~89-91%) vs spatial CV; only OLS/SDM/SDEM coefficients are reported for interpretation, no SHAP/PDP for the ML models.", "confidence": "high"}6