Deep-dive articles on machine learning, AI engineering, and production data science
A multi-objective optimization framework using Lexicographic Particle Swarm Optimization (PSO) and Stacking-based ANN forecasting for active and reactive power flow in smart microgrids under the Moroccan Range Tariff System.
A hybrid deep learning approach combining CNN and BiLSTM, optimized by a Genetic Algorithm, to forecast Morocco's quarterly real GDP under expanding-window out-of-sample conditions.
Learn how Random Forest works under the hood, why it prevents overfitting, and how to build one using Scikit-Learn in Python.
Avoid these common pitfalls in ML: from data leakage during preprocessing to evaluation on the wrong metrics.
Discover the most powerful techniques for engineering features that boost model performance on tabular datasets.
A structured framework to select the right algorithm based on data size, type, explainability, and latency requirements.
Learn about Accuracy, Precision, Recall, F1-Score, and ROC-AUC, and when to use each for model evaluation.
Understand the bias-variance tradeoff and learn the main techniques to prevent overfitting in your machine learning models.
Build clean, production-ready machine learning workflows using Scikit-Learn Pipeline and ColumnTransformer.
An overview of Explainable AI (XAI) techniques like SHAP and LIME to make black-box models transparent and interpretable.
Learn how to use Optuna for efficient, Bayesian-optimized hyperparameter searches in your machine learning pipelines.
Discover how combining multiple weak models leads to a strong model using bagging, boosting, and stacking techniques.
I build custom ML models, RAG chatbots, data pipelines, and production APIs — from analysis to deployment.