⚓ ML system for submarine threat detection using SVM to classify sonar signals as mines or rocks. Features GridSearchCV optimization, sklearn pipelines, and comprehensive evaluation metrics for underwater anomaly recognition.

Training Higgs Dataset with Keras - https://doi.org/10.5281/zenodo.13133945

Predicting heart disease using the UCI dataset with Logistic Regression, KNN, and Random Forest. Models are evaluated with accuracy, precision, recall, F1, and ROC-AUC. A voting ensemble is tested to improve performance, and a Chi-Squared test identifies key predictive features.

Research Paper Grap Modeling

End-to-end ML pipeline for UCI Heart Disease classification. Includes leak-safe preprocessing, baseline + Random Forest + HistGradientBoosting, val-tuned thresholds, and CI that generates a downloadable reports artifact. Best model (HGB) hits F1=0.872, Acc=0.891 on the held-out test set

Phishing Website Detection — a modern Streamlit web app that identifies phishing URLs using a transparent, rule-based engine with instant analysis, detailed reasoning, and a clean UI. Includes a legacy ML pipeline (Random Forest) and data tools for the UCI Phishing Websites dataset.

In diesem Projekt entwickle ich eine vereinfachte „Mini-Schufa“ mithilfe des Machine-Learning-Modells Random Forest. Ziel ist es, die Kreditwürdigkeit eines Nutzers anhand seiner Eingaben einzuschätzen

Student Performance Analysis using Python, SQL & Power BI.

This repository showcases my foundational knowledge in machine learning concepts, based from the Machine Learning series by TechWithTim. It serves as my requirement for aspiring members of the PUP Hygears Programming Team, demonstrating my ability to apply theoretical knowledge to practical implementations.

Automated ML pipeline for the UCI Bank Marketing dataset: ETL, Optuna-based AutoML, model evaluation, MLflow logging, pytest tests, Docker, and CI/CD.

Dashboard de análisis descriptivo del sistema de bicicletas compartidas de Washington D.C. (UCI ML Repo): Python, pandas, matplotlib sobre 731 días y 16 variables sin valores faltantes.

Marketing funnel and conversion performance analysis using the UCI Bank Marketing Campaign Dataset. Includes funnel drop-off analysis, channel performance, dashboard visuals, and actionable growth recommendations.

A complete ML system to predict student academic performance using UCI dataset. Includes regression (grade prediction) + classification (pass/fail), sklearn Pipelines, GridSearchCV, XGBoost, SVM, KNN, and a live Streamlit web app.

Early Detection of Coronary Heart Disease Using Machine Learning - QM640 Capstone (Walsh College, Fall 2025)

Production-grade retail analytics platform — ETL pipeline, RFM segmentation, sales forecasting, FastAPI, MLOps monitoring, and BI dashboards. Built on the UCI Online Retail dataset.

End-to-end heart disease classification pipeline built with Apache SparkML, MLflow experiment tracking, and hyperparameter tuning via CrossValidator.

An end-to-end customer intelligence system built on Supabase Postgres and Streamlit, using SQL analytics and interactive dashboards to analyze customer behavior, segmentation, churn risk, and lifetime value from the Online Retail II dataset.

End-to-end customer segmentation and lifecycle analysis using transactional e-commerce data. Includes RFM analysis, clustering, cohort-based retention insights, and business-driven recommendations.

Classical NLP and ML on ~215k patient drug reviews (UCI): text-based sentiment classification on the dataset's own test split (0.95 ROC-AUC, well above baseline) and per-condition complaint mining via weighted log-odds. Reusable package with a CLI, tests, and reproducible results.

Exercise from freeCodeCamp to practice data analysis with Python and Pandas on the 1994 U.S. Census dataset.