Questions tagged [loss]
599 questions
0
votes
1 answer
Custom loss function with Keras to penalise more negative prediction
I understand that mse will treat both actual - predict, and predict - actual the same way. I want to write a custom loss function such that
the penalty of predict > actual is more than actual > predict
Say I will have 2x more penalty for being…
Hairy Ass
- 190
- 2
- 10
0
votes
0 answers
Tool to measure packet loss for UDP protocol
I am trying to measure packet loss for a UDP video streaming. While streaming the video from source to destination, I run wireshark in the background to capture the traffic. But wireshark doesn't help me measure the packet loss. Is there any tool…
ST94
- 35
- 8
0
votes
1 answer
Plot MSE over epochs when the loss function is a customized function
I use a customzied loss function and would like to plot the MSE within epochs (I use Keras Library). This is the code I use to fit my neural network and save the history.
model.compile(loss =new_loss2, metrics=['mse'], optimizer=opt)
hist =…
Hajar Elhammouti
- 103
- 2
- 10
0
votes
1 answer
Assign values to the tensors in keras while defining a new loss function
I am trying to define a loss function in Keras
def rmseApprox(y_true, y_pred):
dum = y_pred
dum[y_pred>=0]=1.1
dum[y_pred<0]=1
return k.abs(K.mean(y_true - dum*y_pred), axis=-1)
which increase the positive values by a factor of 1.1…
Reza Sabzi
- 1
- 2
0
votes
1 answer
Autoencoder loss and accuracy on a simple binary data
I'm trying to understand and improve the loss and accuracy of the variational autoencoder. I filled the autoencoder with a simple binary data:
data1 = np.array([0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0,
1, 0, 0, 0, 0,…
MarioZ
- 320
- 4
- 17
0
votes
0 answers
Adding a second loss mask in tensorflow
I'm using google's seq2seq library and based on some computations that I do between the predictions and the input I would like to zero out some of the losses for certain time steps (not for padding).
What I do is basically to go through each batch…
lehar
- 3
- 4
0
votes
2 answers
cassandra cluster ping loss
We have 3 nodes Cassandra cluster. 2 nodes are in EU (seeds) and 1 node is in US. Sometimes we have network problems - ping loss between nodes in EU and US. Effect after network problems is that some data is missing on US node. We need to start…
Anton Antonov
- 31
- 4
0
votes
0 answers
Overfitting Deep Neural Network with Regression
I build a neural network with regression for predict a prime of insurance data. My Loss function take a very high value ( like 123000) and decrease unitl 30000; my accuracy remain constant at 1. I'm in a situation of overfitting?
In Traing set I…
jjgasse
- 319
- 2
- 5
- 16
0
votes
0 answers
Keras custom loss function that computes the slope
How do I write a custom loss function of keras that computes the slope of y_pred and y_true and calculates the difference between the two slopes?
Márcia Baptista
- 1
- 5
0
votes
1 answer
[MXNet]Periodic Loss Value when training with "step" learning rate policy
When training deep CNN, a common way is to use SGD with momentum with a "step" learning rate policy (e.g. learning rate set to be 0.1,0.01,0.001.. at different stages of training).But I encounter an unexpected phenomenon when training with this…
user2854015
- 21
- 2
0
votes
1 answer
Custom external loss metric for Gradient Optimizer?
I have an external function which takes y and y_prediction (in matrix format), and computes a metric which depicts how good or bad the prediction actually is.
Unfortunately the metric is no simple y - ypred or confusion matrix, but still very…
Rikku Porta
- 307
- 1
- 4
- 18
0
votes
2 answers
Custom weighted loss in cntk
I like to calculate weighted error below:
def calc_err(pred, targets, weights) :
nClass = np.size(pred, axis=0)
Is = [1.0 for i in range(nClass)]
nonTargets = C.minus(Is, targets)
wrongPred = C.minus(Is, pred)
wColumn =…
Jason Ruan
- 55
- 5
0
votes
1 answer
Very low Error but a low Accuracy in a keras RNN
More of a theoretical question than anything. If I have a near zero cross entropy loss in a binary classification where the last layer is a softmax and the input layer is an LSTM, does it make sense that the accuracy tops out at 54% on the train…
a1letterword
- 307
- 1
- 4
- 16
0
votes
1 answer
python svm function with huber loss
I need a svm classifier of python with huber loss function. But its default loss function is hinge loss. Do you know how can I assign loss function to python svm?
svc = svm.SVC(kernel='linear', C=1, gamma=1).fit(data, label)
user3104352
- 1,100
- 1
- 16
- 34
0
votes
1 answer
Tensorflow Multiple Input Loss Function
I am trying to implement a CNN in Tensorflow (quite similar architecture to VGG), which then splits into two branches after the first fully connected layer. It follows this paper: https://arxiv.org/abs/1612.01697
Each of the two branches of the…
Alan Dolhasz
- 1
- 2