censored-regression

Regression Models for predicting Loss Given Default (LGD) by accounting for the limits of the LGD variable, typically 0% to 100%. In a default scenario, The models predict the expected loss for a loan, where the response variable (LGD) is bounded by total recovery (0) and total loss (1).

This repository contains the R codes used for the simulation analyses in the manuscript titled "Censored Panel Quantile Regression with Fixed Effects via an Asymmetric Link Function" by Fulden Komuryakan & Selahattin Guris.

How much gold did Tibia's The Great Expedition remove from the economy? A censored-likelihood analysis that uses each world's published pass/fail status as data.

Can a food's ingredient list predict its measured acrylamide? A pre-registered test against FDA survey data. Answer: 9.8% RMSE reduction, needed 20%. FAIL.

National abundance trend analysis of England’s river macroinvertebrates (1990–2024). Implements a custom R pipeline using Arrow for big data and seasonal interval-censored GAMMs (mgcv) to reconcile legacy categorical records with modern numeric counts.