Researcher @CISPA | Previously Research Engineer @Cambridge_Uni, @Bitdefender 🇷🇴
Germany
Joined August 2019
- Tweets115
- Following300
- Followers158
- Likes3.8K
Bogdan Cebere retweeted
Join Inria PreMeDICal in Montpellier for a 6-month internship working on the @opendp_org Library! 🇫🇷
💻 Combine theory + code
📈 Path to PhD/Eng roles
📩 Send your CV + start date
More Details: researchers.lille.inria.fr/a…
Bogdan Cebere retweeted
The best photo of May: well done, @NicusorDanRO! Long live Romania, long live Romania in Europe, long live Europe!
Bogdan Cebere retweeted
The iconic Unirii Square Fountains in Bucharest light up in the colours of Europe to celebrate our hard fought victory over fascism! Long live România! Long live Europe! Slava Ukraini!
#romania #elections
Bogdan Cebere retweeted
I need Spotify Wrapped for PDFs:
- you were in the top 0.05% of downloaders
- you were in the bottom 0.2% of readers
Bogdan Cebere retweeted
🔍 Looking for better detection models for Domain Generation Algorithms (DGA)?
🚀 In our new @RAID_Conference paper, we review a some assumptions that can limit the impact of the DGA detection models and discuss some improvements that can lead to more robust detection models.
Bogdan Cebere retweeted
🔐 Interested in Encrypted Traffic Fingerprinting?
🚀 In our new @RAID_Conference paper, we revisit the problem with a focus on the network protocols underneath. This overlooked perspective leads to interesting insights for better attacks and defenses.
🔍 Looking for better detection models for Domain Generation Algorithms (DGA)?
🚀 In our new @RAID_Conference paper, we review a some assumptions that can limit the impact of the DGA detection models and discuss some improvements that can lead to more robust detection models.
Work done with @Jonathan Lasse Bennet Flueren, @silvi_sebastian, @push_pnx and @chrossow.
🎯 Check out the paper and code below:
Paper: dl.acm.org/doi/10.1145/36788…
Code: github.com/bcebere/Guideline…
#dga #dga_detection #malware_analysis #machine_learning
🔐 Interested in Encrypted Traffic Fingerprinting?
🚀 In our new @RAID_Conference paper, we revisit the problem with a focus on the network protocols underneath. This overlooked perspective leads to interesting insights for better attacks and defenses.
Work done with @chrossow at @CISPA.
🎯 Check the work below:
Paper: dl.acm.org/doi/10.1145/36788…
Code: github.com/bcebere/Understan…
#web_fingerprinting #tls #encrypted_traffic_analysis #machine_learning
Awesome book by @m_ou_se , with a ton of important lessons and clear examples. The concurrency primitives are so much clearer now 😁
🔍 Looking for the optimal survival analysis model?
🚀 AutoPrognosis automatically selects a suitable pipeline for time-to-events tasks. Learn how 👇
Tutorial: colab.research.google.com/dr…
Library: github.com/vanderschaarlab/a…
Synthcity offers several differentially-private algorithms for generating synthetic tabular data.
Check out how to control the synthetic privacy level and the quality tier. 👇
colab.research.google.com/dr…
#syntheticdata #tabulardata #privacy
🔍 Synthcity offers a wide range of metrics for testing tabular data quality, downstream performance, statistical fidelity, and privacy.
#syntheticdata #tabulardata #privacy
Learn how to benchmark synthetic generators 👉colab.research.google.com/dr…
Tired of generating synthetic corgis❓🐶
🌟 Check out Synthcity, a library for generating and benchmarking synthetic tabular data. github.com/vanderschaarlab/s…
#syntheticdata #tabulardata #privacy
1/4
🔍 Synthcity supports benchmarking multiple algorithms, testing data quality, downstream performance, statistical fidelity, and privacy metrics.
🌀 Give it a try:
- Library: github.com/vanderschaarlab/s…
- Tutorial: colab.research.google.com/dr…
- Docs: synthcity.readthedocs.io/
Bogdan Cebere retweeted
OpenMined Research is accepting applications!
If you are excited to create novel research in #ML, #PPML, #privacy, #fairness, #accountability, or #transparency, please fill out [0] by 15th of December!
Questions? Send @TCebere a quick DM!
[0]: forms.gle/VLv2wStLHZkRLssm8
