首页 文章

使用Keras ValueError的多类分类：检查目标时出错：

提问于
浏览 次
-1

我正在尝试 Build 一个模型来预测房屋的损坏 . 我正在使用Keras .

在'damage_grade'栏中预测5个值在1到5之间 . 数字越大，房屋可以获得的伤害越大 .

另外我想提一下我是Keras的初学者，这是我在Keras的第一个模特 . 我试图通过Keras documentation的帮助来做到这一点 .

我的代码是：

X_train = rtrain_df.drop("damage_grade", axis=1) 
Y_train = rtrain_df["damage_grade"] 
X_test = rtest_df.drop("building_id", axis=1).copy() 
X_train.shape, Y_train.shape, X_test.shape

import keras 
from keras.models import Sequential 
from keras.layers import Dense, Dropout, Activation 
from keras.optimizers import SGD

model = Sequential() 
model.add(Dense(64, activation='relu', input_dim=46)) #there are 46 feature in my dataset to be trained 
model.add(Dropout(0.5)) 
model.add(Dense(64, activation='relu')) 
model.add(Dropout(0.5)) 
model.add(Dense(10, activation='softmax'))

model.compile(optimizer='rmsprop', loss='categorical_crossentropy', metrics=['accuracy'])

model.fit(X_train, Y_train, epochs=20, batch_size=128)

尝试拟合模型时，会出现以下错误：

ValueError：检查目标时出错：期望dense_6有形状（10，）但得到的形状为数组（1，）

有大约60万条记录需要培训

python machine-learning neural-network keras deep-learning

1 回答

  • 1

    您的代码中存在一些错误：

    • 你在最后一个Dense图层中给出了10 . 它必须等于要预测的值的数量，即5 .

    • 您必须将Y_train转换为具有5个分类要素的分类数组（'damage_grade'，从0到4） .

    以下是更正后的代码：

    X_train = rtrain_df.drop("damage_grade", axis=1) 
Y_train = rtrain_df["damage_grade"] 
X_test = rtest_df.drop("building_id", axis=1).copy() 
X_train.shape, Y_train.shape, X_test.shape

import keras 
from keras.models import Sequential 
from keras.layers import Dense, Dropout, Activation 
from keras.optimizers import SGD

from keras.utils import np_utils
Y_train_cat = np_utils.to_categorical(Y_train) # converts into 5 categorical features

model = Sequential() 
model.add(Dense(64, activation='relu', input_dim=46))
model.add(Dropout(0.5)) 
model.add(Dense(64, activation='relu')) 
model.add(Dropout(0.5)) 
model.add(Dense(5, activation='softmax')) 

# last Dense layer is the output layer that'll produce the probabilities for the 5 
# outputs

model.compile(optimizer='rmsprop', loss='categorical_crossentropy', metrics=['accuracy'])

model.fit(X_train, Y_train_cat, epochs=20, batch_size=128)

import numpy as np

predictions = model.predict(X_test)
result = np.argmax(predictions,axis=1) # sets the output with max probability to 1
    回复于

相关问题