인프런 커뮤니티 질문&답변

예시문제 작업형2를 시험환경에서 lgb 모델로 머신러닝을 실시하였는데, 'random_state' 파라미터 관련 경고 메세지가 뜨는데, lgb 모델에서는 'random_state' 를 사용할 수 없는 건가요?

[코딩내용]

import pandas as pd

X_test = pd.read_csv("data/X_test.csv")

X_train = pd.read_csv("data/X_train.csv")

y_train = pd.read_csv("data/y_train.csv")

pd.set_option('display.max_columns', 100)

pd.options.display.float_format = '{:.2f}'.format

# print(X_train.shape, X_test.shape, y_train.shape)

# print(X_train.head(3))

# print(X_test.head(3))

# print(X_train.info())

# print(X_train.describe())

# print(X_train.describe(include='object'))

# print(X_test.describe(include='object'))

# print(y_train.head(3))

# print(y_train['gender'].value_counts())

# print(X_train.isnull().sum())

# print(X_test.isnull().sum())

# print(y_train.isnull().sum())

# print(X_train['환불금액'].mean())

# print(X_test['환불금액'].mean())

X_train['환불금액'] = X_train['환불금액'].fillna(0)

X_test['환불금액'] = X_test['환불금액'].fillna(0)

# print(X_train.isnull().sum())

# print(X_test.isnull().sum())

X_train = X_train.drop('cust_id', axis=1)

X_test_id = X_test.pop('cust_id')

# print(X_test.head(3))

cols = X_train.select_dtypes(exclude='object').columns

# print(cols)

from sklearn.preprocessing import RobustScaler

scaler = RobustScaler()

X_train[cols] = scaler.fit_transform(X_train[cols])

X_test[cols] = scaler.transform(X_test[cols])

# print(X_train.head(3))

cols = X_train.select_dtypes(include='object').columns

# print(cols)

from sklearn.preprocessing import LabelEncoder

for col in cols :

le = LabelEncoder()

X_train[col] = le.fit_transform(X_train[col])

X_test[col] = le.transform(X_test[col])

# print(X_train.head(3))

from sklearn.model_selection import train_test_split

X_tr, X_val, y_tr, y_val = train_test_split(X_train, y_train['gender'], test_size=0.15, random_state=2022)

# print(X_tr.shape, X_val.shape, y_tr.shape, y_val.shape)

from sklearn.metrics import roc_auc_score

import lightgbm as lgb

model = lgb.LGBMClassifier(ramdom_state=2022, max_depth=5, n_estimators=600, learning_rate=0.01)

model.fit(X_tr, y_tr)

pred = model.predict_proba(X_val)

# print(pred[:10])

print(roc_auc_score(y_val, pred[ : , 1]))

# 0.6153810060060059

# max_depth=5 : 0.6353541041041042

# n_estimators=600, learning_rate=0.01 : 0.647366116116116

[경고 메세지]

[LightGBM] [Warning] Unknown parameter: ramdom_state