ProphitBet is a Machine Learning Soccer Bet prediction application. It analyzes the form of teams, computes match statistics and predicts the outcomes of a match using Advanced Machine Learning (ML) methods. The supported algorithms in this application are Neural Networks, Random Forests & Ensembl Models.

[ICDE'20] ⚖️ A general, efficient ensemble framework for imbalanced classification. | 泛用，高效，鲁棒的类别不平衡学习框架

[NeurIPS’20] ⚖️ Build powerful ensemble class-imbalanced learning models via meta-knowledge-powered resampler. | 设计元知识驱动的采样器解决类别不平衡问题

Bank customers churn dashboard with predictions from several machine learning models.

Experimental implementations of several (over/under)-sampling techniques not yet available in the imbalanced-learn library.

Testing different supervised machine learning algorithms to predict credit risk

Conception and deployment of a credit-scoring model, API and interactive dashboard

Credit risk analysis using scikit-learn and imbalanced-learn.

Loan Default Prediction: Predict loan default risk using classification models with feature engineering, model training, and evaluation metrics.

A full stack classification machine learning project.

Identify credit card risk using Machine Learning algorithms

Data visualization of the NYC restaurant data, and data analysis to gauge if a restaurant located in a high-income area receives a higher health inspection grade. Uses Python (Pandas, Scikit-learn, Imbalanced-learn), PostgreSQL, SQLAlchemy, Tableau, JavaScript (Plotly.js library), HTML, CSS, and Bootstrap.

Onco-Logic is a comprehensive, multi-modal decision support ecosystem designed to transform cancer care by unifying fragmented patient data. The suite leverages advanced AI and machine learning to provide clinicians and researchers with a holistic understanding of each patient's disease, enabling a new frontier in precision oncology.

Credit Card Fraud Detection using Machine Learning – A classification project that detects fraudulent credit card transactions using supervised learning, with data preprocessing, handling class imbalance, and model evaluation (ROC-AUC, Precision, Recall, F1-score).

Demonstrating how changes in input image resolution affect the algorithm's output

Engineered a predictive ML pipeline to classify online shoppers’ purchase intent and segment customer types – leveraging SMOTE to address class imbalance, applying mRMR for feature selection, and training multiple scikit-learn classifiers and K-Means clustering to drive revenue-boosting insights.

Predict credit risk using a variety of Resampling Models and algorithms.

Final project for the end of the course in collaboration with Alessandro Zanzi.

Data cleaning, preprocessing, and class balancing of the Palmer Penguins dataset using Python (pandas, seaborn, scikit-learn, imbalanced-learn). Includes handling missing values, outliers, encoding, visualization, and SMOTE.