Автоматический классификатор объектов железнодорожного транспорта

Credit Fraud Detection of a highly imbalanced dataset of 280k transactions. Multiple ML algorithms(LogisticReg, ShallowNeuralNetwork, RandomForest, SVM, GradientBoosting) are compared for prediction purposes.

The aim of this project is to determine which classification technique produces the best results when applied to the task of determining credit riskiness.

This repository will compare the performance of different classification algorithms on various imbalanced datasets using multiple balancing techniques.

Supervised Machine Learning and Credit Risk

Prática de Árvore de Decisão

Imbalanced data commonly exist in real world, especially in anomaly-detection tasks. Handling imbalanced data is important to the tasks, otherwise the predictions are biased towards the majority class. RandomUnderSampler, ClusterCentroids, CondensedNearestNeighbour, and etc. are useful undersampling tools to remove data for majority classes.

Repo contains scripts to perform data analysis on structure data. It also provides a comparison of various ML algorithms at different stages of data preparation.

Credit Card Fraud Detection Using Machine Learning

Credit Card Fraud detection based on anonymized data using multiple classification algorithms

Multi-label classification project

Predicting the ability of a borrower to pay back the loan through Traditional Machine Learning Models and comparing to Ensembling Methods

Multi-label classification approaches on the Yeast dataset

Detecting Frauds in Online Transactions using Anamoly Detection Techniques Such as Over Sampling and Under-Sampling as the ratio of Frauds is less than 0.00005 thus, simply applying Classification Algorithm may result in Overfitting