Final year student from Telecom ParisTech and HEC Paris, I'm a data science junkie and I'm looking to join a data team. So either you have a killer data startup or you want to build a data-driven recommendation product, feel free to contact me. I DO : ■ Machine learning prototyping ■ Mathematical modeling ■ Data visualization ■ Big Data - Fast Data --- I : ■ Love programming : getting to know new stacks, libraries, and tools is not only a job. It's a crazy hobby ■ Am Interested in data science, NLP, applied machine learning on social networks. ■ Am interested in front-end web technologies ■ Am eager to collaborate to the open-source community --- I PRACTICE: ■ Python. Main libraries: Pandas, SciPy stack, Scikit-learn, Flask, PyMongo, Bokeh, PySpark, Tweepy... Experience: > 2 years. Projects: scripting, scraping data from internet, building web servers and websites backends, managing databases, visualizing data, accessing various APIs in stream mode. ■ Web design: JavaScript, Angular.js, D3.js, HTML, CSS, Flask (Python MicroFramework), Twitter Bootstrap. ■ Database management: MongoDB (noSQL), SQL ■ Apache Spark (Two edX certifications with highest scores) ■ ElasticSearch, Kibana, LogStash ■ OPP: Java. ■ GPU programming : C, Cuda, PyCuda

Data Scientist @ This internship has been a great opportunity in which I got a hands on experience in Data Science. The main purpose was to extract and visualize valuable data out of Orange and Dailymotion log files: Key tools have been involved : ■ GPGPU : General-Purpose Processing on Graphics Processing Units In other words, the idea is to take advantage of GPU highly advanced parallel computing to optimize large database queries in an efficient and remarkably fast fashion. ■ General data aggregation and statistical analyses: R, ElastiSearch + Kibana, Python (Pandas API) ■ Interactive and responsive data visualization : HTML 5 + Bootstrap + d3.js + dc.js + crossilter.js ■ Distributed file systems: Hadoop (MapReduce + HDFS) ■ Processing on streaming data​ : Apache Spark (Python APIs) ■ Use of Twitter streaming API From March 2015 to August 2015 (6 months) Intern in Machine Learning and applied mathematics @ The main goal of the internship has been to prototype an anomaly and fraud detection framework to be integrated in an existing Amadeus web product as an independent and fully autonomous plug-in. For that purpose, a machine learning (ML) approach has been conducted to model frauds as suspicious and anomalous events. The defined road-map has been set as follow: ✔ Evaluation of the fraud impacts on the airline industry. ✔ State-of-the-art and review of many machine learning techniques used in anomaly detection and fraud detection (SVM, SOM, Neural Nets, Nearest Neighbors, ...) ✔ Analysis of the application log files: feature extraction, modeling and storage. ✔ Training phase: A user-centric approach has been followed. For each user, we build a User Behavior Model (UBM) based on the set of previous connections (actions, timestamps, devices, IP addresses, locations etc...) . this statistical model being a description of a normal behavior and a training record. Techniques: Markov Models, Gaussian Mixtures, Jensen-Shannon metrics, Connected Components Clusterings, Dominant Sets clustering ✔ Test phase: For each user, we consider a new test session. We compute its anomaly score by defining a specific metric w.r.t the UBM, then we output a anomaly score. An anomaly score lies in the unit interval and quantifies how anomalous a session is. Techniques: Data mining aggregation operators, Multi-Criteria Decision Analysis, Ordered Weighted-Averaging Operators (OWA), Weighted OWA (WOWA) operators ✔ Data visualization and web interface using HTML5, CSS3, JavaScript (JQurey), MongoDB for the Front-End and Python+Flask for the Back-End. ✔ Success in outlying anomalous behaviors. ✔ Big interest from top management directors. ✔ Participation to several internal Amadeus events to defend this project (see distinctions.) From September 2014 to February 2015 (6 months) Blue-collar intern @ ■ Review of the Tunisian mobile network architecture. ■ Evaluation of telecommunication frauds. From July 2013 to September 2013 (3 months)

Master's degree, Digital Management @ HEC Paris From 2015 to 2016 Master of Science (MSc), Computer sciences @ Telecom ParisTech From 2012 to 2016 Master of Science (MSc), Multimedia communications - Web engineering, GPA : 3.6/4 @ EURECOM From 2013 to 2014 PCSI - PSI*, Mathematics, Physics, Engineering science, Entrance to Telecom-paristech. @ Lycée Pierre de Fermat From 2010 to 2012 Baccalauréat, Mathematics, 18.97/20 | highest honors @ Lycée Pilote de l'Ariana From 2006 to 2010 Ahmed BESBES is skilled in: Machine Learning, Python, Java, Apache Spark, C, Matlab, Data Mining, Web Development, Linux, Software Engineering, MongoDB, C++, CSS, LaTeX, HTML5