Uses Logistic Regression and various machine learning techniques to train and evaluate models with imbalanced classes applied to identify the creditworthiness of borrowers.

A generative model 4 fraudulent transaction synthesis and detection

LightGBM for handling label-imbalanced data with focal and weighted loss functions in binary and multiclass classification

official code repository for TMLR paper "Online Continual Learning via Logit Adjusted Softmax"

A Python library built in Rust for implementing the Hellinger distance splitting criteria in a Random Forest Classifier to address imbalanced data. Work in progress.

This project focuses on predicting customer churn in a telecom company, a critical business metric due to the high cost of acquiring new customers compared to retaining existing ones.

A spam detection model built to handle imbalanced data using small pipelines. This project walks through text preprocessing, model tuning, and performance evaluation with ROC-AUC curves and classification reports, focusing on practical steps like using XGBoost and TFIDF for spam classification.

RCSMOTE: Range-Controlled Synthetic Minority Over-sampling Technique for handling the class imbalance problem

Code repository for the online course Machine Learning with Imbalanced Data

Predict the probability of default for each user id in risk modeling

Predictors for Blood-Brain Barrier Permeability with resampling strategies based on B3DB database.

imFTP: Deep Imbalance Learning via Fuzzy Transition and Prototypical Learning (imFTP, Information Sciences 2024)

웹 광고 클릭률 예측 AI 경진대회, DACON (2024.05.07 ~ 2024.06.03)

Detect robot traffic in an e-commercial website

Research on machine learning, deep learning, and ensemble methods in imbalanced fraud and anomaly detection scenarios.

🛠️ Class-imbalanced Ensemble Learning Toolbox. | 类别不平衡/长尾机器学习库

Churn Analysis of Telecom company,Through meticulous data analysis and predictive modeling, we uncover patterns, trends, and potential churn triggers, empowering telecom companies to proactively mitigate customer attrition. Our mission is to equip industry stakeholders with actionable intelligence, enabling them to optimize retention strategies.

Many algorithms for imbalanced data support binary and multiclass classification only. This approach is made for mulit-label classification (aka multi-target classification). 🌻

This project was completed as part of the CIT 650 "Intro To Big Data" course at Nile University.