-
![default](https://user-images.githubusercontent.com/2363295/38863247-9d3db98a-4269-11e8-9575-2cde39caa74d.png)
![default](https://user-images.githubusercontent.com/2363295/38863286-b8529902-4269-11e…
-
## Keyword: differential privacy
### State-of-the-Art Approaches to Enhancing Privacy Preservation of Machine Learning Datasets: A Survey
- **Authors:** Chaoyu Zhang
- **Subjects:** Cryptography an…
-
The following changes are to be implemented from #2 :
1. Obtain a sample from the function to model as a vector instead of a single number.
2. Use the vectors as batches for training.
Note that the…
-
微博内容精选
-
貼吧活動:(請查閱 [SARS-CoV-2 Timeline by 2020.02.21](https://github.com/agorahub/_meta/blob/agoran/theagora/sari/Memorandum_2020-02-21_SARS-CoV-2-Timeline_Nathan.pdf?raw=true), by Nathan :cloud: )
- Colla…
-
Example: https://mila.quebec/en/publications/
It would be nice to reuse the same code as in the Mila website. Not sure if that's 'easily' possible via RTD
-
I could be misreading NN widget help file [1], but is it possible to specify a single-layer NN - i.e. no hidden layers? If I enter 0 for the number of hidden layers neurons, it gives an error (screens…
-
-
import jax
import jax.numpy as jnp
from emlp.reps import V,Scalar
from emlp.groups import SO
import numpy as np
W =V(SO(3))
rep = 2*W
P = (rep>>Scalar).equivariant_projector()
applyP = lambd…
-
I can't remember where, but there was a recent PR offering cosmetic (e.g. PEP8) fixes, and we usually reject these PRs as causing merge conflicts with existing open PRs. @amueller wanted to know exact…