Hassan CHOUHAD

Developing micro-service for predicting container dweltime and estimating fuel consumption. ACHIEVEMENTS: - Data preprocessing to clean, transform, and prepare data for analysis. - Feature engineering to extract relevant information and improve model performance. - Model's hyperparameter optimization and maximize its predictive capabilities. - ETL Using SQL and MongoDB for efficient data collection, storage, and retrieval. - Creating modular python scripts to respect DevOps and MLOps concept. - Deploy application on Kubernetes using Docker containers and scale securely on SSH servers. - Strong problem-solving skills and data-driven approach to tackle complex business challenges. - Proficiency in visualizing and interpreting results through effective data visualization techniques. - Predictive modeling based on machine learning (XGBoost and neural networks) to accurately estimate container dwelltime. - Collecting, preprocessing, and analyzing historical data to train prediction models. - developing an EDA for statistical Analysis, and validating a Proof of Concept. - Thorough validation and evaluation of models to ensure reliable and high-quality results. - Close collaboration with the development team to integrate the Flask API into a larger system. - Comprehensive documentation of the Flask API, including endpoints, input and output parameters, and usage examples. - Developing a POC to generate a meeting minute using OpenAI, and ChatGPT. - Proactive in staying updated on the latest advancements in AI methodologies and data science. Compétences : MLflow · Jupyter · Scaleway · MongoDB · VScode · Linux · Git · Python (langage de programmation) · Docker · Scikit-learn · TensorFlow · Kubernetes · Data Preprocessing · MySQL · SQL · Deep learning · Data Analysis · Machine learning · NoSQL · Data Science

Towards online metrology for proactive quality control in smart manufacturing. - Detect and classify defects on 4K pixel images using CNN, transfer learning for increasing the model performance to reach 70% accuracy on test data. - Defect localization on images using MaskRCNN to perform an instance segmentation. - Predicting the surface roughness of machined parts with Neural Networks and ensemble methods (random forest, GBoost, XGBoost) using scikit-learn, visual-code, Jupyter notebook and other data science libraries. - Applying a model’s comparison in terms of model’s errors and the processing time. - Dimensionality reduction of the dataset using the PCA method and finding correlation matrix. - Analysis and post-processing of raw data using Matlab and MountainsMap. - Producing ideas for data collection, data analysis, and Machine Learning model building for classification and prediction. - Coding algorithms and knowing when to apply them. - Performing an EDA with jupyter-notebook for a statistics analysis, data analytics and data visualization. - Write progress reports. - Authoring scientific papers. - Exhaustive scientific research on Calculus, linear algebra, and the mathematical theories behind the machine learning algorithms. - Presenting the results and progress of the dissertation together with the supervision team and the STIL company’s team. - Presentation of achievements and training of employees of industrial partner STIL Marposs. - Bibliographic research using Mendeley for references management. TOOLS: Python, MATLAB, Anaconda, Visual Studio, TensorFlow-GPU, scikit-learn, TensorFlow, PowerShell, ImageJ, office 365, OpenCV, MountainsMap.

Building and Design of an access control system based on professional skills. - Construction of the system's hardware with Raspberry-Pi. - Development of Python code for access management. - Exploitation of the APIs from a PHP web application developed by the project manager. - Creation of a MySQL database. - Writing a technical report. TOOLS: Python, MySQL, SQL, Linux Raspbian, Atom, and RasperryPi.

Development of a web application for bus fleet management. - Analysis and adaptation of two Open-Source applications (IoT platform Thingsboard and Traccar). - UML design for modeling the functions and use cases. - Design of database models with Sails.JS, and MongoDB after trying the SpringBoot. - Writing of project report. TOOLS: Sails.JS, MongoDB, Git, Apache tomcat, AngularJS, Node.js, Office.