TeX 62.3%
Python 37.6%
1\begin{table}[htbp]2\centering\scriptsize3\caption{Task A — Point accuracy of future sale-price forecasts by horizon band (months after the origin)}\label{tab:scores_a}4\begin{threeparttable}5\begin{tabular}{lrrrrrr}6\toprule7Model & 1-3: MAE & 1-3: MASE & 4-6: MAE & 4-6: MASE & 7-12: MAE & 7-12: MASE \\8\midrule9Naive: municipal median & 34.96 & 1.00 & 35.27 & 1.00 & 35.29 & 1.00 \\10Hedonic OLS & 23.24*** & 0.66 & 23.80*** & 0.67 & 24.44*** & 0.69 \\11Ridge & 23.24*** & 0.66 & 23.80*** & 0.67 & 24.44*** & 0.69 \\12Elastic net & 23.19*** & 0.66 & 23.73*** & 0.67 & 24.38*** & 0.69 \\13SVR (subsample) & 22.99*** & 0.66 & 23.42*** & 0.66 & 23.65*** & 0.67 \\14Random forest & 22.04*** & 0.63 & 22.59*** & 0.64 & 23.36*** & 0.66 \\15XGBoost & 21.36*** & 0.61 & 22.27*** & 0.63 & 23.02*** & 0.65 \\16LightGBM & 21.21***$^{\dagger}$ & 0.61 & 22.04***$^{\dagger}$ & 0.62 & 22.70***$^{\dagger}$ & 0.64 \\17Quantile MLP & 26.14*** & 0.75 & 25.97*** & 0.74 & 25.98 & 0.74 \\18HSHN (no pooling) & 22.34*** & 0.64 & 23.66*** & 0.67 & 24.82*** & 0.70 \\19HSHN (full pooling) & 22.55*** & 0.65 & 23.98*** & 0.68 & 25.19*** & 0.71 \\20HSHN (partial pooling) & 22.36*** & 0.64 & 23.68*** & 0.67 & 24.83*** & 0.70 \\21\bottomrule22\end{tabular}23\begin{tablenotes}\footnotesize\item Target: log sale price. MAE in log points ×100; MASE relative to the naive reference. Stars: Diebold--Mariano test of equal squared-error loss against the reference on per-origin mean losses (* 10\%, ** 5\%, *** 1\%); $\dagger$ = in the 90\% model confidence set of the horizon.\end{tablenotes}24\end{threeparttable}25\end{table}26