Master's student in Big Data & Data Science with a background in Mathematics. I build end-to-end machine learning pipelines across NLP, Computer Vision, and predictive modeling. Based in Sousse, Tunisia.
- M.Sc. Big Data & Data Science — Horizon University, Sousse
- B.Sc. Mathematics — Higher School of Science and Technology, Sousse
- Strong mathematical foundation applied to deep learning and statistical modeling
- Experience building and deploying ML systems from data preprocessing to production APIs
- Competitive chess player — strategic thinking and pattern recognition transfer directly to ML
Languages: Python, R, Java, SQL
Deep Learning: PyTorch, TensorFlow/Keras, CNNs, Transfer Learning, GRU/BiGRU
NLP: Sequence modeling, Machine Translation, HuggingFace Transformers
Computer Vision: YOLOv8, Object Detection, Image Segmentation, TensorRT, DeepStream
Databases: SQL, MongoDB, Redis, Cassandra
Tools: Git, FastAPI, Jupyter, VS Code, RStudio, RAPIDS
| Project | Description | Tech |
|---|---|---|
| Sentiment Analysis: BiLSTM vs BERT | Binary sentiment classification on IMDb reviews — BiLSTM from scratch vs fine-tuned BERT. F1 improved from 0.86 to 0.94 | PyTorch, HuggingFace, Gradio |
| NLP Machine Translation | French to English translation — LSTM Seq2Seq vs Transformer (MarianMT). BLEU score improved from 0.23 to 39.38 | PyTorch, HuggingFace, Gradio |
| Smart Agriculture — Plant Disease Detection & Segmentation | Fine-tuned YOLOv8 for plant disease detection and segmentation on PlantDoc dataset. mAP50: 0.513, Dice Score: 0.684. Deployed with FastAPI + Streamlit | YOLOv8, FastAPI, Streamlit, PyTorch |
- DeepLearning.AI & Stanford — Machine Learning Specialization (Supervised Learning, Deep Learning, Unsupervised Learning, Reinforcement Learning)
- IBM — Data Science Professional Certificate (Python, SQL, ML, Data Visualization)
- NVIDIA — Deep Learning, Real-Time Video AI, Accelerated Data Science, Transformer-Based NLP
- LinkedIn: linkedin.com/in/muhammedattia
- Email: attiamuhammed1@gmail.com
Always building. Always learning.