Naoaki ONO
Reputation: 91
TensorFlow - optimization with normalization constraints
Is there any implementation using TensorFlow to optimize a vector under a constraint such as |w|^2==1 ?
Upvotes: 2
Views: 1022
Answers (1)
Rafał Józefowicz
Reputation: 6235
You can add a soft constraint to your loss: some_constant * (norm(w)- 1)^2 but, as far as I know, there are no functionalities specifically for constrained optimization.
Upvotes: 3
Related Questions
- Tensorflow: how to minimize under constraints
- Tensorflow multi-dimension argmax
- Implementing Normalization Inside Tensorflow Model
- min-max normalization of a tensor in PyTorch
- How to normalize tensor inside custom loss function in TensorFlow?
- tensorflow - normalizing vectors
- Batch Normalization - Tensorflow
- data normalization in tensorflow
- Construct a layer with tensorflow to normalize a matrix/tensor
- Implementing batch normalization with tensorflow