Cost-Sensitive Learning / ReSampling / Weighting / Thresholding / BorderlineSMOTE / AdaCost / etc.

Value-driven and cost-sensitive analysis for scikit-learn

Theano implementation of Cost-Sensitive Deep Neural Networks

A complete end-to-end fraud detection system for financial transactions, featuring data pipelines, cost-sensitive ML modeling, explainability with SHAP, threshold optimization, batch scoring, and an interactive Streamlit dashboard. Designed to simulate real-world fintech fraud-risk workflows.

This repo contains implementation of advanced ML techniques. Includes model ensembles, cost-sensitive learning and dealing with class imbalance.

Pytorch implementation for paper 'BANNER: A Cost-Sensitive Contextualized Model for Bangla Named Entity Recognition'

Advanced Machine Learning Algorithms including Cost-Sensitive Learning, Class Imbalances, Multi-Label Data, Multi-Instance Learning, Active Learning, Multi-Relational Data Mining, Interpretability in Python using Scikit-Learn.

A hands-on lab showing how “improving” a single metric (AUC/accuracy/F1) can worsen real-world outcomes. Includes metric audits, slice checks, cost-sensitive evaluation, threshold tuning, and decision policies you can defend, so dashboards don’t quietly ship bad decisions.

A python implementation of a genetic algorithm based approach for cost sensitive learning

End-to-end diabetes risk prediction pipeline (Pima): EDA → feature engineering → calibration + cost-aware threshold → deployable artifacts.

A genetic algorithm based approach for cost sensitive learning, in which the misclassification cost is considered together with the cost of feature extraction.

A machine learning project addressing credit card fraud detection using imbalanced datasets. Utilizes techniques like cost-sensitive learning, SMOTE, and ensemble models for high precision and accuracy, emphasizing robust performance despite challenging data distributions.

Official code for our paper - "Melanoma classification from dermatoscopy images using knowledge distillation for highly imbalanced data".

Deep Cost-sensitive Kernel Machine Model - PAKDD 2020

Implementation of cost sensitive KNN algorithm described in Qin, et al, 2013

Solution to the Data Mining Cup 2019 competition

Cost-aware credit card fraud detection pipeline: time-based split, probability calibration, and business-aligned threshold tuning (AUPRC-first).

Predicting whether an African country will be in recession or not with advanced machine learning techniques involving class imbalance, cost-sensitive learning and explainable machine learning

Worked on detecting illicit transactions in the Ethereum Transactions dataset by increasing our dataset size, and with little tolerance to missing fraudulent transactions.

Code repository for the book "Imbalanced Data: Myths, Mistakes and Modern Solutions".