我想使用KerasCLassifier来解决多类分类问题。变量y的值是经过one-hot编码的，例如：
0 1 01 0 01 0 0这是我的代码：
from keras.models import Sequentialfrom keras.layers import Densefrom keras.wrappers.scikit_learn import KerasClassifier# Function to create model, required for KerasClassifierdef create_model(optimizer='rmsprop', init='glorot_uniform'): # create model model = Sequential() model.add(Dense(2048, input_dim=X_train.shape[1], kernel_initializer=init, activation='relu')) model.add(Dense(512, kernel_initializer=init, activation='relu')) model.add(Dense(y_train_onehot.shape[1], kernel_initializer=init, activation='softmax')) # Compile model model.compile(loss='categorical_crossentropy', optimizer=optimizer, metrics=['accuracy']) return model# create modelmodel = KerasClassifier(build_fn=create_model, class_weight="balanced", verbose=2)# grid search epochs, batch size and optimizeroptimizers = ['rmsprop', 'adam']epochs = [10, 50]batches = [5, 10, 20]init = ['glorot_uniform', 'normal', 'uniform']param_grid = dict(optimizer=optimizers, epochs=epochs, batch_size=batches, init=init)grid = model_selection.GridSearchCV(estimator=model, param_grid=param_grid, scoring='accuracy')grid_result = grid.fit(X_train], y_train_onehot)当我运行最后一行代码时，在10个epoch后抛出了以下错误：
/opt/conda/lib/python3.6/site-packages/sklearn/metrics/classification.py in accuracy_score(y_true, y_pred, normalize, sample_weight) 174 175 # Compute accuracy for each possible representation –> 176 y_type, y_true, y_pred = _check_targets(y_true, y_pred) 177 check_consistent_length(y_true, y_pred, sample_weight) 178 if y_type.startswith(‘multilabel’):
/opt/conda/lib/python3.6/site-packages/sklearn/metrics/classification.py in _check_targets(y_true, y_pred) 79 if len(y_type) > 1: 80 raise ValueError(“Classification metrics can’t handle a mix of {0} ” —> 81 “and {1} targets”.format(type_true, type_pred)) 82 83 # We can’t have more than one value on y_type => The set is no more needed
ValueError: Classification metrics can’t handle a mix of multilabel-indicator and binary targets
当我将accuracy替换为categorical_accuracy或balanced_accuracy时，我无法编译模型。
回答：
这是一个工作演示：
import numpy as npfrom sklearn.model_selection import GridSearchCVfrom keras.models import Sequentialfrom keras.layers import Densefrom keras.wrappers.scikit_learn import KerasClassifierN = 100X_train = np.random.rand(N, 4)Y_train = np.random.choice([0,1,2], N, p=[.5, .3, .2])# Function to create model, required for KerasClassifierdef create_model(optimizer='rmsprop', init='glorot_uniform'): # create model model = Sequential() model.add(Dense(2048, input_dim=X_train.shape[1], kernel_initializer=init, activation='relu')) model.add(Dense(512, kernel_initializer=init, activation='relu')) model.add(Dense(len(np.unique(Y_train)), kernel_initializer=init, activation='softmax')) # Compile model model.compile(loss='sparse_categorical_crossentropy', optimizer=optimizer, metrics=['sparse_categorical_accuracy']) return model# create modelmodel = KerasClassifier(build_fn=create_model, class_weight="balanced", verbose=2)# grid search epochs, batch size and optimizeroptimizers = ['rmsprop', 'adam']epochs = [10, 50]batches = [5, 10, 20]init = ['glorot_uniform', 'normal', 'uniform']param_grid = dict(optimizer=optimizers, epochs=epochs, batch_size=batches, init=init)grid = GridSearchCV(estimator=model, param_grid=param_grid, scoring='accuracy')grid_result = grid.fit(X_train, Y_train)PS 请注意sparse_categorical_*损失函数和度量标准的使用。
