Python 在keras中实现自定义目标函数
我正在尝试实施我自己的成本函数,具体如下: 现在我知道这个问题在这个网站上被问了好几次,我读到的答案通常如下:Python 在keras中实现自定义目标函数,python,machine-learning,keras,loss-function,Python,Machine Learning,Keras,Loss Function,我正在尝试实施我自己的成本函数,具体如下: 现在我知道这个问题在这个网站上被问了好几次,我读到的答案通常如下: def custom_objective(y_true, y_pred): .... return L 人们似乎总是使用y\u true和y\u pred然后说你只需要编译模型model.compile(loss=custom\u objective),然后从那里开始。没有人真正提到代码中的某个地方y\u true=something和y\u pred=something。这是我必
def custom_objective(y_true, y_pred):
....
return L
人们似乎总是使用y\u true
和y\u pred
然后说你只需要编译模型model.compile(loss=custom\u objective)
,然后从那里开始。没有人真正提到代码中的某个地方y\u true=something
和y\u pred=something
。这是我必须在模型中指定的吗
我的代码
不确定我是否正确使用了.predict()
从正在训练的模型中获取运行预测:
params = {'lr': 0.0001,
'batch_size': 30,
'epochs': 400,
'dropout': 0.2,
'optimizer': 'adam',
'losses': 'avg_partial_likelihood',
'activation':'relu',
'last_activation': 'linear'}
def model(x_train, y_train, x_val, y_val):
l2_reg = 0.4
kernel_init ='he_uniform'
bias_init ='he_uniform'
layers=[20, 20, 1]
model = Sequential()
# layer 1
model.add(Dense(layers[0], input_dim=x_train.shape[1],
W_regularizer=l2(l2_reg),
kernel_initializer=kernel_init,
bias_initializer=bias_init))
model.add(BatchNormalization(axis=-1, momentum=momentum, center=True))
model.add(Activation(params['activation']))
model.add(Dropout(params['dropout']))
# layer 2+
for layer in range(0, len(layers)-1):
model.add(Dense(layers[layer+1], W_regularizer=l2(l2_reg),
kernel_initializer=kernel_init,
bias_initializer=bias_init))
model.add(BatchNormalization(axis=-1, momentum=momentum, center=True))
model.add(Activation(params['activation']))
model.add(Dropout(params['dropout']))
# Last layer
model.add(Dense(layers[-1], activation=params['last_activation'],
kernel_initializer=kernel_init,
bias_initializer=bias_init))
model.compile(loss=params['losses'],
optimizer=keras.optimizers.adam(lr=params['lr']),
metrics=['accuracy'])
history = model.fit(x_train, y_train,
validation_data=[x_val, y_val],
batch_size=params['batch_size'],
epochs=params['epochs'],
verbose=1)
y_pred = model.predict(x_train, batch_size=params['batch_size'])
history_dict = history.history
model_output = {'model':model,
'history_dict':history_dict,
'log_risk':y_pred}
return model_output
然后创建模型:
model(x_train, y_train, x_val, y_val)
迄今为止我的目标函数
“日志风险”将是y\u true
,而x\u train
将用于计算y\u pred
:
def avg_partial_likelihood(x_train, log_risk):
from lifelines import CoxPHFitter
cph = CoxPHFitter()
cph.fit(x_train, duration_col='survival_fu_combine', event_col='death',
show_progress=False)
# obtain exp(hx)
cph_output = pd.DataFrame(cph.summary).T
# summing hazard ratio
hazard_ratio_sum = cph_output.iloc[1,].sum()
# -log(sum(exp(hxj)))
neg_log_sum = -np.log(hazard_ratio_sum)
# sum of positive events (death==1)
sum_noncensored_events = (x_train.death==1).sum()
# neg_likelihood
neg_likelihood = -(log_risk + neg_log_sum)/sum_noncensored_events
return neg_likelihood
尝试运行时出错
AttributeError回溯(最近一次调用)
在()
---->1个模型(x_-train,y_-train,x_-val,y_-val)
模型中(x_-train,y_-train,x_-val,y_-val)
45模型编译(损失=平均部分可能性,
46 optimizer=keras.optimizers.adam(lr=params['lr']),
--->47个指标=[‘准确度’])
48
49历史=模型拟合(x_系列、y_系列、,
编译中的~\Anaconda3\lib\site packages\keras\engine\training.py(self、优化器、损耗、度量、损耗权重、样本权重模式、加权度量、目标张量、**kwargs)
331带有K.name_作用域(self.output_names[i]+''u loss'):
332输出损耗=加权损耗,
-->333样品(重量,面罩)
334如果透镜(自输出)>1:
335自度量张量追加(输出损失)
~\Anaconda3\lib\site packages\keras\engine\training\u utils.py加权(y\u true,y\u pred,weights,mask)
401 """
402#分数_数组的ndim>=2
-->403分数数组=fn(y_真,y_pred)
404如果掩码不是无:
405#将遮罩投射到floatX上,以避免float64向上投射
平均部分可能性(x列、对数风险)
27
28 cph.fit(x列、持续时间、生存时间、死亡时间),
--->29显示(进度=错误)
30
31#获得经验(hx)
~\Anaconda3\lib\site packages\lifelines\fitters\coxph\u fitter.py in fit(self、df、duration\u col、event\u col、show\u progress、initial\u beta、strata、step\u size、weights\u col)
90 """
91
--->92 df=df.copy()
93
94#按时分拣
AttributeError:“Tensor”对象没有属性“copy”
没有人真正提到代码中的某个地方
y\u true=something
和y\u pred=something
他们没有提到这一点,因为您不需要这样做!实际上,在每个过程结束时(即在一个批上向前传播),Keras使用模型的真实标签和预测为该过程提供
y_-true
和y_-pred
。因此,您根本不需要在您的模型中定义y_-true
和y_-pred
。只需使用后端函数定义损失函数(即,将Keras导入后端的定义为K
)而且一切都会很好地工作(并且永远不要在丢失函数中使用numpy)。要想了解更多信息,请查看in-Keras,看看它们是如何实现的。这是一个可用后端函数的列表(可能不完整)。非常感谢!
AttributeError Traceback (most recent call last)
<ipython-input-26-cf0236299ad5> in <module>()
----> 1 model(x_train, y_train, x_val, y_val)
<ipython-input-25-d0f9409c831a> in model(x_train, y_train, x_val, y_val)
45 model.compile(loss=avg_partial_likelihood,
46 optimizer=keras.optimizers.adam(lr=params['lr']),
---> 47 metrics=['accuracy'])
48
49 history = model.fit(x_train, y_train,
~\Anaconda3\lib\site-packages\keras\engine\training.py in compile(self, optimizer, loss, metrics, loss_weights, sample_weight_mode, weighted_metrics, target_tensors, **kwargs)
331 with K.name_scope(self.output_names[i] + '_loss'):
332 output_loss = weighted_loss(y_true, y_pred,
--> 333 sample_weight, mask)
334 if len(self.outputs) > 1:
335 self.metrics_tensors.append(output_loss)
~\Anaconda3\lib\site-packages\keras\engine\training_utils.py in weighted(y_true, y_pred, weights, mask)
401 """
402 # score_array has ndim >= 2
--> 403 score_array = fn(y_true, y_pred)
404 if mask is not None:
405 # Cast the mask to floatX to avoid float64 upcasting in Theano
<ipython-input-23-ed57799a1f9d> in avg_partial_likelihood(x_train, log_risk)
27
28 cph.fit(x_train, duration_col='survival_fu_combine', event_col='death',
---> 29 show_progress=False)
30
31 # obtain exp(hx)
~\Anaconda3\lib\site-packages\lifelines\fitters\coxph_fitter.py in fit(self, df, duration_col, event_col, show_progress, initial_beta, strata, step_size, weights_col)
90 """
91
---> 92 df = df.copy()
93
94 # Sort on time
AttributeError: 'Tensor' object has no attribute 'copy'