D

data-science

Analisi della transizione energetica nell'UE27 (2000–2030) con Python, Power BI e Machine Learning. Tre indici sintetici originali (ITE, ICP, RGI) calcolati su dati Our World in Data e proiettati al 2030 tramite regressione lineare.

Progetto IFTS Data Analysis & AI - SIAM1838

A practical, linear-algebra-first introduction to data science.

Supervised learning pipeline for rare event operational failure prediction, integrating leakage resistant preprocessing, class weighted modeling, precision recall threshold calibration, ROC AUC benchmarking, and permutation based feature importance to analyze production stress drivers.

Projet d’analyse du comportement des utilisateurs Instagram à partir d’un large dataset synthétique (plus d’un million d’utilisateurs).

Le projet explore :

Analyse exploratoire des données (EDA) Prédictions de variables comportementales (stress, âge, clics publicitaires, revenu) Identification de profils utilisateurs avec clustering (KMeans) Interprétation des résultats statistiques en lien avec la vie réelle

Méthodologie : visualisation, corrélations, sélection de modèles, entraînement, évaluation et interprétation.

Technologies : Python, Pandas, NumPy, Matplotlib, Scikit-learn, Jupyter.

Projet réalisé dans le cadre de la formation Développeur en Intelligence Artificielle (Simplon).

API FastAPI permettant d’identifier des espèces de pingouins à partir de données tabulaires optionnelles et/ou d’une image.

L’API renvoie une liste d’espèces probables avec leurs probabilités et inclut :

Classification Machine Learning (RF, KNN, LR) Fusion multimodale (tabulaire + image) Journalisation des requêtes/réponses (SQLite) Interface web pour consulter les logs Interface web pour tester les prédictions Notebooks pour exploration et entraînement

Projet réalisé dans le cadre de la formation Développeur en Intelligence Artificielle (Simplon).

This repo will have all resources, labs, data which I use/d on Kaggle Network

Tool-driven analytics agent using LLM orchestration, schema-based tools, and a semantic data dictionary.

A set of data science and machine learning projects exploring various datasets — a place to test ideas, models, and analytical approaches.

Comparação entre Vision Transformers e Métodos Clássicos de Visão Computacional na Segmentação de Exsudatos Lipídicos em Imagens de Retinopatia Diabética. Trabalho de Conclusão de Curso para obtenção de título de bacharel em Engenharia Elétrica na Universidade Federal do Ceará.

Built a complete machine learning pipeline in SparkML using the Adult Census dataset (~48k rows, 14 features). Implemented data preprocessing, feature encoding, cross-validation, and model training with Logistic Regression and Random Forest. Evaluated models with metrics such as AUC and F1-score. Reflected on scalability trade-offs and optimizations in distributed ML.

This repo is a mix of several data science tools. There is a mix of web-scraping of data that is then cleaned, and used to analyze the property market in malta, using prediction models, visualisations and statistical analysis.

There is also visualisations for chess data from the 1980's till 2021. Moreover, there is twitter data, which is then stored in the neo4J nosql dbms.

No data is presented in the git, only the results. Code with the data can be found at: https://drive.google.com/file/d/15EQnRtsngDsFDD_A7g4N1fwCuXI0f_Xi/view?usp=sharing

[Uc3m] The Classic ghost-eating game

Capstone 1: Caterpillar Tube Pricing Prediction & Categorization. Capstone 2:Pipeline Multi-Leak Classification.

A replacement for patsy better suitable for fully automated use

Home credicard competition in kaggle.

Code and notebooks related to my successfully defended final project for my Master's of Science in Data Science degree at Lewis University, entitled “Utility Pole Measurements and Feature Analysis Using Computer Vision and Photogrammetry”