我在 Pandas 数据框中有一个训练集,我将此数据框传递到model.fit()
with df.values
。以下是有关 df 的一些信息:
df.values.shape
# (981, 5)
df.values[0]
# array([163, 0.6, 83, 0.52,
# array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
# 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
# 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
# 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
# 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
# 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
# 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
# 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
# 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
# 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
# 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
# 0, 0, 0, 0, 0, 0, 0])], dtype=object)
正如您所看到的,df 中的行包含 5 列,其中 4 列包含数值(int 或 float),一列包含表示某些分类数据的热编码数组。我正在创建我的 keras 模型,如下所示:
model = keras.Sequential([
keras.layers.Dense(1024, activation=tf.nn.relu, kernel_initializer=init_orth, bias_initializer=init_0),
keras.layers.Dense(512, activation=tf.nn.relu, kernel_initializer=init_orth, bias_initializer=init_0),
keras.layers.Dense(256, activation=tf.nn.relu, kernel_initializer=init_orth, bias_initializer=init_0),
keras.layers.Dense(128, activation=tf.nn.relu, kernel_initializer=init_orth, bias_initializer=init_0),
keras.layers.Dense(64, activation=tf.nn.relu, kernel_initializer=init_orth, bias_initializer=init_0),
keras.layers.Dense(1, activation=tf.nn.sigmoid)
])
opt = keras.optimizers.Adam(lr=0.001, beta_1=0.9, beta_2=0.999, epsilon=None, decay=0.0, amsgrad=True)
model.compile(optimizer=opt,
loss='binary_crossentropy',
metrics=['accuracy'])
model.fit(df.values, df_labels.values, epochs=10, batch_size=32, verbose=0)
df_labels.values
只是一个由 0 和 1 组成的一维数组。所以我相信最后确实需要一个 Dense(1) sigmoid 层,以及“binary_crossentropy”损失。
如果我只传递数值数据,这个模型效果很好。但是一旦我引入热编码(分类数据),我就会收到此错误:
ValueError Traceback (most recent call last)
<ipython-input-91-b5e6232b375f> in <module>
42 #trn_values = df_training_set.values[:,:,len(df_training_set.columns)]
43 #trn_cat = df_trn_wtid.values.reshape(-1, 1)
---> 44 model.fit(df_training_set.values, df_training_labels.values, epochs=10, batch_size=32, verbose=0)
45
46 #test_loss, test_acc = model.evaluate(df_test_set.values, df_test_labels.values)
~\Anaconda3\lib\site-packages\keras\engine\training.py in fit(self, x, y, batch_size, epochs, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, initial_epoch, steps_per_epoch, validation_steps, **kwargs)
1037 initial_epoch=initial_epoch,
1038 steps_per_epoch=steps_per_epoch,
-> 1039 validation_steps=validation_steps)
1040
1041 def evaluate(self, x=None, y=None,
~\Anaconda3\lib\site-packages\keras\engine\training_arrays.py in fit_loop(model, f, ins, out_labels, batch_size, epochs, verbose, callbacks, val_f, val_ins, shuffle, callback_metrics, initial_epoch, steps_per_epoch, validation_steps)
197 ins_batch[i] = ins_batch[i].toarray()
198
--> 199 outs = f(ins_batch)
200 outs = to_list(outs)
201 for l, o in zip(out_labels, outs):
~\Anaconda3\lib\site-packages\keras\backend\tensorflow_backend.py in __call__(self, inputs)
2713 return self._legacy_call(inputs)
2714
-> 2715 return self._call(inputs)
2716 else:
2717 if py_any(is_tensor(x) for x in inputs):
~\Anaconda3\lib\site-packages\keras\backend\tensorflow_backend.py in _call(self, inputs)
2653 array_vals.append(
2654 np.asarray(value,
-> 2655 dtype=tf.as_dtype(tensor.dtype).as_numpy_dtype))
2656 if self.feed_dict:
2657 for key in sorted(self.feed_dict.keys()):
~\Anaconda3\lib\site-packages\numpy\core\numeric.py in asarray(a, dtype, order)
536
537 """
--> 538 return array(a, dtype, copy=False, order=order)
539
540
ValueError: setting an array element with a sequence.
请不要建议将 one_hot 数组中的每个值扩展到它们自己的列中。此示例是我的数据集的精简版本,其中包含 6-8 个分类列,其中一些 one_hots 是大小超过 5000 的数组。所以这对我来说不是一个可行的解决方案。我希望改进我的顺序模型(或彻底修改 keras 模型),以便处理分类数据和数值数据。
请记住,训练标签是 0/1 值的一维数组。我需要用数值/分类训练集来预测一组结果,我不能从数值数据中得到一组预测,并从分类数据中得到一组预测。