ngrislain

9 ms·

61.
Advances in GenAI change the game for differential privacy(sarus.tech)
7 points by ngrislain 3y ago|0 comments
62.
by ngrislain 4y ago
nl
63.
by ngrislain 5y ago
You are absolutely right, we are leveraging many open-source bricks to build our product, so that they can be reviewed, mainly: - https://github.com/google/differential-privacy (for basic mechanisms and PLD accounting)
64.
by ngrislain 5y ago
Sarus would typically fit in organizations that legitimately collect personal data and can take decisions based on these data. In these cases you don't want (and most of the time cannot legally) let anyone in the organisation have acce
65.
Distributed ML with Dask and Kubernetes on GCP(medium.com)
1 points by ngrislain 5y ago|1 comments
66.
by ngrislain 5y ago
At Sarus Technologies, we use dask to test many machine learning models and tune their hyper-parameters in parallel. It's actually quite simple to deploy and scale a dask cluster, thanks to kubernetes and public clouds. If you are inte
67.
Estimating a cumulative distribution function with differential privacy(medium.com)
1 points by ngrislain 6y ago|0 comments
68.
An interactive model of election forecast to play with martingale property(observablehq.com)
1 points by ngrislain 6y ago|1 comments
69.
by ngrislain 6y ago
Recently, as I was following the 🇺🇸 US elections 🇺🇸, I got puzzled by the fact that days after the elections happened some betting odds of Biden winning the elections were still below 90%, while FiveThirtyEight published a probability o…
70.
A Not-So-Secret Ballot(medium.com)
2 points by ngrislain 6y ago|2 comments
71.
by ngrislain 6y ago
A Bayesian perspective on how differential privacy could maintain voter anonymity
72.
Clean implementations in TF2 of recent generative models – by Sarus Tech(github.com)
2 points by ngrislain 6y ago|1 comments
73.
by ngrislain 6y ago
# Sarus published models Sarus implementation of classical ML models. The models are implemented using the Keras API of tensorflow 2. Vizualization are implemented and can be seen in tensorboard. The required packages are managed with pipen