test predict이 먹히지 않습니다.

# 최종 제출 파일 (lightGBM)
pred = lg.predict(test)
submit=pd.DataFrame({'pred':pred})
submit.to_csv("result.csv",index=False)

아래처럼 feature가 맞지 않아서 에러가뜹니다. 강의 그대로 했는데 왜그러는 걸까요??
---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
/tmp/ipython-input-50-2529582812.py in <cell line: 0>()
      1 # 최종 제출 파일 (lightGBM)
----> 2 pred = lg.predict(test)
      3 submit=pd.DataFrame({'pred':pred})
      4 submit.to_csv("result.csv",index=False)

2 frames
/usr/local/lib/python3.11/dist-packages/lightgbm/sklearn.py in predict(self, X, raw_score, start_iteration, num_iteration, pred_leaf, pred_contrib, validate_features, **kwargs)
   1008         n_features = X.shape[1]
   1009         if self._n_features != n_features:
-> 1010             raise ValueError(
   1011                 "Number of features of the model must "
   1012                 f"match the input. Model n_features_ is {self._n_features} and "

ValueError: Number of features of the model must match the input. Model n_features_ is 28 and input n_features is 8

combined = pd.concat([train, test], axis=0)
combined = pd.get_dummies(combined)

train = combined[:len(train)]
test = combined[len(train):]