Mohamed DYAB - WIZBII Mohamed DYAB a publié son profil professionnel sur WIZBII. M D

Mohamed DYAB

Data Scientist

34 ans • Paris

Résumé

Data Scientist at Dolead (French AdTech startup in Paris). Worked at CNRS, INRIA, Valeo, co-founded startup. Experience in machine learning, deep neural networks, Python, R, Keras, Scikit-learn, SQL, Spark, Statistical analysis, Agile methodology, Test Driven Development. Went to ENSIMAG, Grenoble Alpes University and Alexandria University.

Compétences

PythonrC++machine learningDeep LearningScikit-LearnKerasSQLSparkLinux EnvironmentShell Scriptinggit

Expériences

Research Intern - Deep Learning

- Développement informatiqueTopic:​ ​ Talking-face​ ​ Detection​ ​ using​ ​ Audio​ ​ and​ ​ Visual​ ​ Modalities​ ​ in​ ​ TV​ ​ Shows.​ ​ Python,​ ​ Keras​ ​ and​ ​ Scikit-Learn​ ​ are​ ​ used.​ ​ Source​ ​ code​ ​ on​ ​ GitHub: https://github.com/dyabmo/plumcot ○ Trained​ ​ convolutional​ ​ neural​ ​ networks​ ​ and​ ​ long​ ​ short-term​ ​ memory​ ​ networks​ ​ (LSTM)​ ​ to​ ​ detect​ ​ talking​ ​ faces​ ​ in​ ​ videos,​ ​ using​ ​ both​ ​ visual​ ​ and​ ​ audio​ ​ modalities. ○ Accomplishments:​​ ​ Showed​ ​ that​ ​ using​ ​ a ​ ​ combination​ ​ of​ ​ visual​ ​ and​ ​ audio​ ​ modalities​ ​ for​ ​ LSTM​ ​ improves​ ​ the​ ​ performance​ ​ over​ ​ using​ ​ either​ ​ modality​ ​ alone.

Data Scientist

- MaintenantDéveloppement informatiqueAnalysis and optimization of search engine advertising campaigns using statistical methods and machine learning techniques. Analyze Key performance indicators, search for areas of improvement, build prototypes and evaluate them, implement the best solution in production. Improve keyword bidding using both classical machine learning (Random Forests etc..) and deep learning techniques (LSTMs)

Research Intern - Machine Learning

- Développement informatiqueTopic:​ ​ GPU​ ​ Execution​ ​ Time​ ​ Prediction​ ​ using​ ​ Machine​ ​ Learning​ ​ and​ ​ Analytical​ ​ Modeling.​ ​ Python,​ ​ Scikit-Learn​ ​ and​ ​ R ​ ​ are​ ​ used.​ ​ Source​ ​ code​ ​ on​ ​ GitHub: https://github.com/marcosamaris/MLvsAMgpuperf ○ Trained​ ​ three​ ​ different​ ​ machine​ ​ learning​ ​ models​ ​ to​ ​ predict​ ​ execution​ ​ time​ ​ of​ ​ applications​ ​ running​ ​ on​ ​ GPUs,​ ​ and​ ​ compared​ ​ them​ ​ with​ ​ an​ ​ analytical​ ​ model. ○ Accomplishments: Showed that machine learning approach provides more flexibility than analytical approach, as some variables needn’t be explicitly calculated. Published​ ​ paper​ ​ at​ ​ IEEE-NCA​ ​ 2016​ ​ conference​.

Research Intern

- Développement informatiqueTopic:​ ​ Using​ ​ Resources​ ​ from​ ​ a ​ ​ Closely-related​ ​ Language​ ​ to​ ​ Develop​ ​ ASR​ ​ For​ ​ a ​ ​ Very​ ​ Under-resourced​ ​ Language:​ ​ A ​ ​ case​ ​ study​ ​ for​ ​ Iban​ ​ language. ○ Trained​ I ​ ban​ ​ acoustic​ ​ model​ ​ using​ ​ only​ ​ Iban​ ​ speech​ ​ (without​ ​ transcript)​ ​ and​ ​ acoustic​ ​ model​ ​ of​ ​ Malay​ ​ language​ ​ (a​ ​ closely-related​ ​ language). ○ Accomplishments:​​ ​ My​ ​ reported​ ​ results​ ​ are​ ​ included​ ​ in​ ​ a ​ ​ paper​ ​ published​ ​ at​ ​ Interspeech​ ​ 2015​ ​ conference.

Software Engineer

- Energie / Matériaux / MécaniqueDeveloped​ ​ standard​ ​ embedded​ ​ software​ ​ components​ ​ in​ ​ C/C++.​ ​ Performed​ ​ unit​ ​ testing,​ ​ integration​ ​ testing​ ​ and​ ​ hardware​ ​ testing. ○ Configured​ ​ AUTOSAR​ ​ (AUTomotive​ ​ Open​ ​ System​ ​ Architecture)​ ​ modules,​ ​ specifically​ ​ microcontroller​ ​ abstraction​ ​ layer,​ ​ and​ ​ testing​ ​ them​ ​ on​ ​ hardware​ ​ boards. ○ Accomplishments: Developed several software components such as SMS manager, eCall manager etc.. from scratch for eCall system (Emergency response to car accidents)​ ​ and​ ​ successfully​ ​ tested​ ​ them​ ​ on​ ​ hardware​ ​ boards.

Cofounder ​ ​ & ​ ​ Verification ​ ​ Engineer

- Développement informatiqueProviding​ ​ digital​ ​ systems​ ​ services:​ ​ digital​ ​ IP​ ​ cores​ ​ and​ ​ verification​ ​ IP​ ​ cores:​ ​ HDMI​ ​ 4.0​ ​ stimuli​ ​ generator​ ​ & ​ ​ LEON3​ ​ processor​ ​ configuration. ○ Accomplishments:​ ​ Received​ ​ a ​ ​ cheque​ ​ from​ ​ one​ ​ customer​ ​ for​ ​ our​ ​ efforts​ ​ in​ ​ completing​ ​ his​ ​ project.

Formations

Ensimag - Université Grenoble Alpes

2015 - 2017 Grenoble, France, GrenobleDéveloppement informatique, Études / Statistiques / Data, Développement informatique

Mes qualités

Entreprenant·e
Curieux·se
Persévérant·e
Réactif·ve

Langues parlées

  • Anglais

    Langue maternelle

  • Français

    Intermédiaire

  • Arabe

    Langue maternelle

Expériences Extra Professionnelles

Organizer

Software​ ​ Freedom​ ​ Day​ ​ 2010
-

​ Contacting ​ ​ the ​ ​ event ​ ​ speakers ​ ​ and ​ ​ welcoming ​ ​ them. ​ ​ Handling ​ ​ event’s ​ ​ logistics.

Organizer

Step​ ​ Up​​ ​ Entrepreneurship​ ​ Program
-

​ Planning ​ ​ and ​ ​ executing ​ ​ social ​ ​ entrepreneurship ​ ​ projects.

Ma présence sur internet

Faire de chaque avenir une réussite.
  • Annuaire emplois
  • Annuaire entreprises
  • Événements
    •