Raphael Azorin

Raphael Azorin

prof_pic.jpg

I am a data scientist with a PhD in Machine Learning from Sorbonne University. I worked at EURECOM lab and Huawei’s Paris Research Center on AI for computer networks.

Prior to that, I held several data positions in companies ranging from start-ups to large corporations. I graduated from PSL, where I was enrolled in work/study programs from BSc to MSc.

Before transitioning to Computer Science, I was in marketing and analytics. Back then, I helped build Markentive, the French leading marketing automation agency (from 5 to 30 employees).

news

Aug 26, 2024 :bookmark: Our paper Fine-grained Attention in Hierarchical Transformers for Tabular Time-series was presented at KDD’24 10th Mining and Learning from Time-Series Workshop.
Jul 20, 2024 :pencil2: I contributed to the Cuckoo hashing Wikipedia page.
Jun 18, 2024 :mortar_board: I successfully defended my PhD thesis on Traffic Representations for Network Measurements.
May 23, 2024 :bar_chart: Attended the IHES workshop on Mathematics for and by Large Language Models.
Mar 28, 2024 :bookmark: Our paper Taming the Elephants: Affordable Flow Length Prediction in the Data Plane will be presented at CoNEXT 2024.
Dec 8, 2023 :round_pushpin: Volunteered at CoNEXT’23 @CNAM in Paris
Dec 6, 2023 :speech_balloon: Presented our poster Memory-Efficient Random Forests in FPGA SmartNICs at CoNEXT’23
Dec 5, 2023 :bookmark: Our paper SPADA: A Sparse Approximate Data Structure Representation for Data Plane Per-Flow Monitoring has been accepted in PACMNET
Aug 28, 2023 :mortar_board: Enrolled at the Mediterranean Machine Learning summer school in Thessaloniki (M2L 2023)
Feb 14, 2023 :bookmark: Our workshop paper “It’s a Match” - A Benchmark of Task Affinity Scores for Joint Learning has been accepted at AAAI’23 2nd International Workshop on Practical Deep Learning.
Dec 9, 2022 :speech_balloon: Our extended abstract Learned Data Structures for Per-Flow Measurements has been accepted at CoNEXT’22 student workshop.
Nov 15, 2022 :bookmark: Our paper Towards a Systematic Multi-Modal Representation Learning for Network Data has been accepted at HotNets’22.
Dec 7, 2021 :speech_balloon: Presented our extended abstract Towards a Generic Deep Learning Pipeline for Traffic Measurements at CoNEXT’21 student workshop
Aug 1, 2021 :round_pushpin: Volunteered at SIGCOMM’21
May 10, 2021 :clipboard: Gave a tutorial on link prediction at PSL executive MSc in Paris Graph Analytics lab.