Generalized Linear Models applied to Malaria, Food Insecurity and TB data

This study applies parametric survival analysis to colon cancer patient data, comparing seven parametric models — Exponential, Weibull, Gamma, LogNormal, LogLogistic, Gompertz, and Generalized Gamma — fitted via maximum likelihood estimation using the flexsurvreg function in R. Model selection is based on AIC and log-likelihood criteria. The Generalized Gamma model is identified as the best-fitting model (AIC = 8212.9), revealing a hump-shaped, non-monotone hazard function that peaks around day 300--365 post-treatment. Key findings include a mean survival time of 5.84 years, a 5-year survival probability of 55%, and a critical high-risk window in the first 12 months following surgery. Life functions including the survival function S(t), hazard function h(t), density f(t), and cumulative distribution function F(t) are estimated and interpreted under the best-fitting model.

Tugas

WordStream: Predictive Text Algorithm Presentation

Bài tập lớn môn Phân tích Big Data &UD trong Kinh tế của nhóm Vũ Minh Thư, sử dụng bộ số liệu số 2

Tugas Ketidakpastian Estimasi

trial

This project conducts a comprehensive survival analysis of a rats dataset to investigate factors influencing survival time. It applies both non-parametric and parametric methods, including Kaplan–Meier estimation, Cox proportional hazards models, and accelerated failure time (AFT) models. The analysis evaluates the effects of treatment and sex on survival while accounting for clustering within litters, and compares different modeling approaches to identify the most appropriate framework for describing survival patterns.

Car MPG Predictor Pitch