对比全特征和去WOE特征对MLP性能的影响

def solve():
    from pyodide.http import open_url
    from io import StringIO
    loans_featured_csv = open_url(____).read()
    import pandas as pd
    from sklearn.neural_network import MLPClassifier
    from sklearn.model_selection import train_test_split
    from sklearn.metrics import roc_auc_score
    from sklearn.preprocessing import StandardScaler
    from io import StringIO
    df = pd.read_csv(StringIO(____))
    all_num=[c for c in df.select_dtypes(____).columns if c not in ['____','____'] and not c.endswith(____)]
    no_woe=[c for c in all_num if not c.endswith(____)]
    y=df['____']
    for label,cols in [(____),(____)]:
        X=df[cols].fillna(____)
        X_tr,X_te,y_tr,y_te=train_test_split(____)
        sc=StandardScaler(); Xs_tr=sc.fit_transform(____); Xs_te=sc.transform(____)
        mlp=MLPClassifier(hidden_layer_sizes=(____,____),max_iter=____,random_state=____)
        mlp.fit(____)
        print(____)