DATA SCIENCE PYTHON PLAYGROUND

Scaled optimisation must return meaningful output units.

MLPRegressor, TransformedTargetRegressor and nested width parameters.

Revisit: Neural regression checkpoint

Keep target transformation inside fitting, compare widths and report inverse-transformed predictions.

This complete script replaces the editor contents. It includes the input setup and the workflow.

import numpy as np
import pandas as pd
df=pd.read_csv('data/ml-learning/ML-X15.csv')
from sklearn.base import clone
from sklearn.model_selection import train_test_split, KFold, cross_validate, cross_val_predict, GridSearchCV
from sklearn.dummy import DummyRegressor
from sklearn.metrics import root_mean_squared_error

import matplotlib.pyplot as plt
import seaborn as sns
X=df[['fixed acidity', 'volatile acidity', 'citric acid', 'residual sugar', 'chlorides', 'free sulfur dioxide', 'total sulfur dioxide', 'density', 'pH', 'sulphates', 'alcohol']]
y=df['quality']
X_train,X_test,y_train,y_test=train_test_split(X,y,test_size=.2,random_state=42)
folds=KFold(n_splits=5,shuffle=True,random_state=42)
candidates={}
grids={}
# Prepare inputs and construct the candidate pipeline.
from sklearn.preprocessing import StandardScaler
preprocessor = StandardScaler()
from sklearn.pipeline import Pipeline
from sklearn.neural_network import MLPRegressor
from sklearn.compose import TransformedTargetRegressor
from sklearn.preprocessing import StandardScaler
network = MLPRegressor(hidden_layer_sizes=(24,), max_iter=800, early_stopping=True, tol=0.001, random_state=42)
model = TransformedTargetRegressor(regressor=network, transformer=StandardScaler())
pipeline = Pipeline([('prepare', preprocessor), ('model', model)])
model = pipeline
candidates['mlp_reg']=model
grids['mlp_reg']={'model__regressor__hidden_layer_sizes': [(16,), (24,)]}
# Validate initial candidates on matching training folds.
initial_results={name:cross_validate(candidate,X_train,y_train,cv=folds,scoring='neg_root_mean_squared_error') for name,candidate in candidates.items()}
# Compare with a simple reference before tuning.
reference_results=cross_validate(DummyRegressor(strategy='mean'),X_train,y_train,cv=folds,scoring='neg_root_mean_squared_error')
rows=[]
selected={}
for name,candidate in candidates.items():
    initial=initial_results[name]
    grid=grids[name]
    if grid:
        search=GridSearchCV(candidate,grid,cv=folds,scoring='neg_root_mean_squared_error',n_jobs=1)
        search.fit(X_train,y_train)
        selected[name]=search.best_estimator_
        validation_score=float(search.best_score_)
        settings=search.best_params_
    else:
        selected[name]=candidate
        validation_score=float(np.mean(initial['test_score']))
        settings='Keep defaults: no production search'
    rows.append({'model':name,'initial_score':float(np.mean(initial['test_score'])),'selected_score':validation_score,'settings':str(settings)})
cv_results=pd.DataFrame(rows).set_index('model')
# This solution nominates by mean training-fold score. Other evidence-based choices can be defensible.
chosen_name=cv_results.selected_score.idxmax()
chosen=selected[chosen_name]
oof_predictions=cross_val_predict(chosen,X_train,y_train,cv=folds)
diagnostic_y=y_train
residuals=pd.DataFrame({'actual':diagnostic_y,'predicted':oof_predictions,'residual':diagnostic_y-oof_predictions})
fig,ax=plt.subplots(figsize=(6,4))
sns.scatterplot(x=residuals.predicted,y=residuals.residual,s=12,ax=ax)
ax.axhline(0,color='black')
ax.set(xlabel='Training-validation prediction',ylabel='Residual',title='Training-only residuals')
fig.tight_layout()
fig.savefig('diagnostic.png',dpi=150,bbox_inches='tight')
final_model=clone(chosen).fit(X_train,y_train)
final_predictions=final_model.predict(X_test)
final_rmse=root_mean_squared_error(y_test,final_predictions)
print('Final RMSE:',final_rmse)
loss_curves={}
for name,candidate in selected.items():
    fitted=candidate.named_steps['model']
    if hasattr(fitted,'regressor_'):
        fitted=fitted.regressor_
    if hasattr(fitted,'loss_curve_'):
        loss_curves[name]=list(fitted.loss_curve_)

import numpy as np
import pandas as pd
df=pd.read_csv('data/ml-learning/ML-X15.csv')

Why these steps exist

• The input is the exact fixed deduplicated 600-row sample. Feature scaling and target transformation are fitted within each neural training fit; width comparison and loss/residual evidence precede final evaluation. Predicting through the wrapper restores quality-score units for RMSE.

Another valid approach

Equivalent ordinary Python is welcome. A defensible candidate or grouping need not match the solution’s choice; objective evidence must remain consistent.