Skip to content
View TarekDjaker's full-sized avatar

Highlights

  • Pro

Block or report TarekDjaker

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
TarekDjaker/README.md

Tarek Djaker | Data Scientist Junior

LinkedIn Email GitHub

M2 Mathématiques & Applications (M2A) — Sorbonne Université. Je transforme des données complexes en décisions business : modélisation fiable (pipelines scikit-learn, CV temporelle), mise en prod légère (FastAPI, Docker) et communication claire (dashboards Power BI/Tableau). Disponible immédiatement en Île-de-France.

Actuellement

  • Portefeuille de projets concrets : énergie (prévision conso/production), mobilité urbaine, assistants RAG.
  • PoC d’early stopping (RKHS/boosting) reproductible.

Je recherche un poste de Data Scientist Junior pour mettre en production des modèles utiles et mesurables.


🧰 Stack

Python (pandas, NumPy, polars, matplotlib) · SQL (CTE, fenêtres) · scikit-learn (Pipeline/ColumnTransformer, GridSearchCV) · XGBoost/LightGBM · time series · Power BI / Tableau · FastAPI · Docker · Git/GitHub · tests/packaging · LangChain/LangGraph (RAG)

🚀 Projets (pinned)

  • eco2mix-forecast — Prévision conso/production électrique (RTE éCO₂mix). API FastAPI /predict, modèle joblib, Dockerfile, dashboard BI. → repo (à publier)
  • velib-availability-analytics — Agrégations spatio-temporelles, heatmaps horaires, stations critiques, rapport d’insights. → repo (à publier)
  • kaggle-titanic-baseline — Pipeline « entretien » (GridSearchCV, courbes d’apprentissage, ROC-AUC). → repo
  • rag-eco2mix-faq — Assistant RAG (retriever, éval faithfulness), LangGraph minimal, démo CLI. → repo (à publier)
  • rkhs-boosting-early-stopping — Expérimentations + notes (règle de discrépance). → repo (à publier)
  • data-challenge-elmy — Différentiel de prix Intraday–Spot (features énergie, protocole d’évaluation). → repo (à publier)

📌 Épingle ces 6 dépôts + ajoute une bannière (cover) & topics cohérents.

Pinned Loading

  1. -Furgal-AI-Early-Stopping -Furgal-AI-Early-Stopping Public

    Jupyter Notebook

  2. LunarLand-Project LunarLand-Project Public

    LunarLand Project

    Jupyter Notebook

  3. Question-Answering-Project- Question-Answering-Project- Public

    Python

  4. racousin/data_science_practice_2024 racousin/data_science_practice_2024 Public

    Jupyter Notebook 1 2

  5. Criteo-project Criteo-project Public template

    Python