PERBANDINGAN REGRESI COX DENGAN JARINGAN SYARAF TIRUAN PADA SIMULASI DATA SURVIVAL; (THE COMPARATION OF COX REGRESSION AND ARTIFICIAL NEURAL NETWORK ON SURVIVAL DATA SIMULATION)

Tahir, Danardono

2015 | Disertasi | FMIPA

This thesis discusses about the comparison of Artificial neural network model and the Cox regression analysis of survival data. Cox regression is a model that is most often used in the analysis of survival data, when we want to know the relation between the time to event with the predictor variables. But while the assumption of this regression is not fulfilled, another model is needed as an alternative. Artificial neural network is a nonparametric regression model proposed as an alternative model because it requires only minimal assumptions. To compare these two models, four different simulation schemes, based on Monte Carlo simulation, were used. then the censoring of data as much as 20%, 40%, 60% and 80%, further analysis of the accuracy of predictions of both models using the concordance index, the accuracy of the results obtained Artificial Neural Network model is better than the Cox regression model accuracy. so that it can be concluded that the Artificial Neural Network model can be used as an alternative model in survival data analysis.

Kata Kunci : N