feat: update model parameters for Random Forest, Gradient Boosting, XGBoost, and LightGBM to enhance performance

app/training/generate_deep_figures.py

@@ -229,11 +229,18 @@ def fig_learning_curve(model, scaler, X, y):
                  fontsize=13, fontweight="bold")
     ax.legend(loc="lower right", framealpha=0.85)
     gap = tr_mean[-1] - val_mean[-1]
-    diagnosis = "düşük varyans (iyi)" if gap < 0.03 else "overfitting işareti"
+    if gap > 0.05:
+        diagnosis = "yüksek varyans — regularizasyon gerekli"
+    elif gap > 0.03:
+        diagnosis = "orta varyans — kabul edilebilir"
+    else:
+        diagnosis = "düşük varyans (iyi)"
     ax.annotate(
-        f"Son gap: {gap:.4f}\n→ {diagnosis}",
-        xy=(0.55, 0.05), xycoords="axes fraction",
-        fontsize=10,
+        f"Train-Val Gap: {gap:.4f}\n→ {diagnosis}\n"
+        f"Not: Tree ensemble train score\n"
+        f"yapısal olarak ~1.0 olur",
+        xy=(0.42, 0.05), xycoords="axes fraction",
+        fontsize=9,
         bbox=dict(boxstyle="round,pad=0.5", facecolor=PALETTE["bg"],
                   edgecolor=PALETTE["primary"], alpha=0.85),
     )

app/training/train_classifier.py

@@ -254,9 +254,9 @@ def _build_candidates() -> list[tuple[str, Any]]:
             "Random Forest",
             RandomForestClassifier(
                 n_estimators=200,
-                max_depth=8,
-                min_samples_leaf=15,
-                min_samples_split=20,
+                max_depth=10,
+                min_samples_leaf=10,
+                min_samples_split=15,
                 class_weight="balanced",
                 random_state=42,
                 n_jobs=-1,
@@ -265,12 +265,12 @@ def _build_candidates() -> list[tuple[str, Any]]:
         (
             "Gradient Boosting",
             GradientBoostingClassifier(
-                n_estimators=150,
-                max_depth=3,
-                learning_rate=0.1,
-                subsample=0.7,
-                min_samples_leaf=20,
-                min_samples_split=30,
+                n_estimators=200,
+                max_depth=4,
+                learning_rate=0.08,
+                subsample=0.75,
+                min_samples_leaf=12,
+                min_samples_split=20,
                 random_state=42,
             ),
         ),
@@ -305,15 +305,15 @@ def _build_candidates() -> list[tuple[str, Any]]:
         candidates.append((
             "XGBoost",
             xgb.XGBClassifier(
-                n_estimators=150,
-                max_depth=3,
-                learning_rate=0.1,
-                subsample=0.7,
-                colsample_bytree=0.7,
-                min_child_weight=10,
-                reg_alpha=0.5,
-                reg_lambda=2.0,
-                gamma=0.3,
+                n_estimators=200,
+                max_depth=4,
+                learning_rate=0.08,
+                subsample=0.75,
+                colsample_bytree=0.75,
+                min_child_weight=8,
+                reg_alpha=0.3,
+                reg_lambda=1.5,
+                gamma=0.2,
                 scale_pos_weight=1.0,
                 eval_metric="logloss",
                 random_state=42,
@@ -325,15 +325,15 @@ def _build_candidates() -> list[tuple[str, Any]]:
         candidates.append((
             "LightGBM",
             lgb.LGBMClassifier(
-                n_estimators=150,
-                max_depth=3,
-                learning_rate=0.1,
-                num_leaves=8,
-                subsample=0.7,
-                colsample_bytree=0.7,
-                min_child_weight=10,
-                reg_alpha=0.5,
-                reg_lambda=2.0,
+                n_estimators=200,
+                max_depth=4,
+                learning_rate=0.08,
+                num_leaves=12,
+                subsample=0.75,
+                colsample_bytree=0.75,
+                min_child_weight=8,
+                reg_alpha=0.3,
+                reg_lambda=1.5,
                 class_weight="balanced",
                 random_state=42,
                 verbose=-1,