I have deep learning and I have questions.
1. Since the optimum weight parameter is the parameter that minimizes loss and val_loss, if val_loss is rampant as shown in the figure below, is it better to pick out the parameter when val_loss falls by chance?
loss and val_loss are defined as follows.
loss: train data answer and prediction error mse
val_loss: test data answer and prediction error mse
2. Why is val_loss so rampage?
It is strange that the deep learning that I have been doing so far has only a pattern in which the loss and val_loss generally fall in minor.
-
Answer # 1
-
Answer # 2
1, I think that the smaller the val_loss, the more reliable the model is (it is better to take out the parameters when the val_loss drops).
2, I can't affirm because there is no code, but it is probably a problem of learning rate.
Increasing the learning rate increases the possibility of divergence instead of faster learning.
If the learning rate is reduced, the learning speed becomes slower at the expense of convergence. -
Answer # 3
1. If val_loss fluctuates greatly, parameter optimization does not work in the first place, so it makes no sense to extract parameters when val_loss is small.
Related articles
- python - inconsistency in sample size of cnn machine learning with keras
- how to apply machine learning (svm) when a list is included in a python explanatory variable (parameter)
- python - i want to get the parameters of the model using state_dict and find the accuracy
- python - how to implement semi-supervised learning on pytorch
- python - about gpu of notebook pc to run machine learning system
- python - [deep learning] i want to display none in the structure of the learner displayed by plot_model
- python - questions about learning with neural network models
- python - error when learning keras neural network
- python - in machine learning learning, the accuracy is 100% and the value of the loss function is almost 0 from the beginning an
- python - transfer learning with tensorflow object detection api is not possible
- python - [machine learning] if the score changes significantly depending on how you divide it, are you overfitting?
- even if i write the code according to 14 of python learning course ii in progate, it is not executed in the same way
- python - about operation parameters in machine learning production
- python - machine learning feature extraction of time series data
- python - about formal parameters in the definition
- python - about the ratio of machine learning training data, validation data, and test data
- python - machine learning training/verification data, test data accuracy rate and adjustment
- python - about learning mnist data with keras
- python 3x - python keras about the shape of learning data
- optimal description of python def function
- [python] i don't know how to solve the error
- python - error that no data is entered in x_train_b
- [python] i don't understand the meaning of the error
- python - input data format when implementing multivariate lstm by keras
- python - unable to interpret the results in the implementation of time series forecasting with keras
- python - probability prediction using multiple time series data
- python - 'numpyndarray' object has no attribute'numpy' error
- python - 'numpyndarray' object has no attribute'append' error
- python - inconsistency in sample size of cnn machine learning with keras
- cross-validation in python
This depends on the situation, and sometimes it's better or worse. But it ’s better to manage the rampage first.
I think that it will not be useful unless we estimate the required performance in advance and learn to meet that performance stably.
If there is little validation data, the range of fluke/out of contact will be larger. After that, when the validation data is not the same as the learning data.
If you are simply unable to learn well, you need to review the model. The optimizer can manage it, but it can also be a problem that cannot be solved in the first place.