Multiverse-of-madness-

Advanced machine learning solution for the ModelForge 2026 competition, featuring data analysis, feature engineering, CatBoost, LightGBM, XGBoost, cross-validation, model optimization, and final prediction generation.

ModelForge 2026 ML Competition

Advanced machine learning solution for the ModelForge 2026 competition.

Overview

This repository contains the end-to-end machine learning workflow developed for the ModelForge 2026 competition, including:

  • Exploratory data analysis
  • Data preprocessing
  • Feature engineering
  • Cross-validation
  • CatBoost
  • LightGBM
  • XGBoost
  • Model comparison and optimization
  • Final prediction generation

Dataset

The competition dataset contains information related to credit card accounts, including:

  • Credit limit
  • Demographic variables
  • Repayment status
  • Bill amounts
  • Previous payment amounts

The target variable is:

default payment next month

Built With

  • jupyter-notebook
Share this project:

Updates