AI & Data Science Research Notes/Research and project notes for AI and Data Science.Evaluating classification systems beyond accuracyA practical framework for precision, recall, F1, calibration, and operational cost.From raw CSV to reproducible feature pipelineHow to document cleaning, leakage controls, lineage, and feature ownership.Responsible deployment checklist for student ML projectsA compact model card, monitoring, rollback, privacy, and fairness checklist.Understanding embeddings through geometric intuitionCosine similarity, vector spaces, clustering, and retrieval with clear examples.Time-series validation without future leakageWalk-forward validation, seasonality, baselines, and honest error measurement.Designing an AI portfolio that proves engineering abilityTurn notebooks into deployable, documented, testable, and useful products.