9.9.0 updated documentation

Author: mathias-von-ottenbreit

API_REFERENCE_FOR_CLASSIFICATION.md

@@ -59,7 +59,7 @@ Specifies a penalty in the range [0.0, 1.0] on terms that are not linear effects
 Specifies a penalty in the range [0.0, 1.0] on interaction terms. A higher value increases model interpretability but can hurt predictiveness. Values outside of the [0.0, 1.0] range are rounded to the nearest boundary within the range.
 
 #### max_terms (default = 0)
-Restricts the maximum number of terms in any of the underlying models trained to ***max_terms***. The default value of 0 means no limit. After the limit is reached, the remaining boosting steps are used to further update the coefficients of already included terms. A potential tuning objective could be to find the lowest positive value of ***max_terms*** that does not increase the prediction error significantly. Low positive values can speed up the training process significantly.
+Restricts the maximum number of terms in any of the underlying models trained to ***max_terms***. The default value of 0 means no limit. After the limit is reached, the remaining boosting steps are used to further update the coefficients of already included terms. An optional tuning objective could be to find the lowest positive value of ***max_terms*** that does not increase the prediction error significantly. Low positive values can speed up the training process significantly. Setting a limit with ***max_terms*** may require a higher learning rate for best results.
 
 
 ## Method: fit(X:npt.ArrayLike, y:List[str], sample_weight:npt.ArrayLike = np.empty(0), X_names:List[str]=[], cv_observations: npt.ArrayLike = np.empty([0, 0]), prioritized_predictors_indexes:List[int]=[], monotonic_constraints:List[int]=[], interaction_constraints:List[List[int]]=[], predictor_learning_rates: List[float] = [], predictor_penalties_for_non_linearity: List[float] = [], predictor_penalties_for_interactions: List[float] = [])

API_REFERENCE_FOR_REGRESSION.md

@@ -127,7 +127,7 @@ Specifies a penalty in the range [0.0, 1.0] on terms that are not linear effects
 Specifies a penalty in the range [0.0, 1.0] on interaction terms. A higher value increases model interpretability but can hurt predictiveness. Values outside of the [0.0, 1.0] range are rounded to the nearest boundary within the range.
 
 #### max_terms (default = 0)
-Restricts the maximum number of terms in any of the underlying models trained to ***max_terms***. The default value of 0 means no limit. After the limit is reached, the remaining boosting steps are used to further update the coefficients of already included terms. A potential tuning objective could be to find the lowest positive value of ***max_terms*** that does not increase the prediction error significantly. Low positive values can speed up the training process significantly.
+Restricts the maximum number of terms in any of the underlying models trained to ***max_terms***. The default value of 0 means no limit. After the limit is reached, the remaining boosting steps are used to further update the coefficients of already included terms. An optional tuning objective could be to find the lowest positive value of ***max_terms*** that does not increase the prediction error significantly. Low positive values can speed up the training process significantly. Setting a limit with ***max_terms*** may require a higher learning rate for best results.
 
 
 ## Method: fit(X:npt.ArrayLike, y:npt.ArrayLike, sample_weight:npt.ArrayLike = np.empty(0), X_names:List[str]=[], cv_observations: npt.ArrayLike = np.empty([0, 0]), prioritized_predictors_indexes:List[int]=[], monotonic_constraints:List[int]=[], group:npt.ArrayLike = np.empty(0), interaction_constraints:List[List[int]]=[], other_data: npt.ArrayLike = np.empty([0, 0]), predictor_learning_rates: List[float] = [], predictor_penalties_for_non_linearity: List[float] = [], predictor_penalties_for_interactions: List[float] = [])

documentation/APLR 9.9.0.pdf

@@ -49,7 +49,7 @@
         verbosity=2,
         m=3000,
         v=0.1,
-        # max_terms=5,  # max terms in each underlying model. Tune this to find a trade-off between interpretability and predictiveness.
+        # max_terms=5,  # Optionally tune this to find a trade-off between interpretability and predictiveness. May require a higher learning rate for best results.
         **params
     )
     model.fit(

examples/train_aplr_classification.py

@@ -40,9 +40,9 @@
     }
 )
 best_model = None
-loss_function = "mse"  # other available loss functions are binomial, poisson, gamma, tweedie, group_mse, mae, quantile, negative_binomial, cauchy, weibull and custom_function.
+loss_function = "mse"  # Other available loss functions are binomial, poisson, gamma, tweedie, group_mse, mae, quantile, negative_binomial, cauchy, weibull and custom_function.
 link_function = (
-    "identity"  # other available link functions are logit, log and custom_function.
+    "identity"  # Other available link functions are logit, log and custom_function.
 )
 for params in param_grid:
     model = APLRRegressor(
@@ -52,7 +52,7 @@
         v=0.1,
         loss_function=loss_function,
         link_function=link_function,
-        # max_terms=10,  # max terms in each underlying model. Tune this to find a trade-off between interpretability and predictiveness.
+        # max_terms=10,  # Optionally tune this to find a trade-off between interpretability and predictiveness. May require a higher learning rate for best results.
         **params
     )
     model.fit(