vighnesh
December 25th, 2007, 07:56 AM
Hi all,
I have developed a black box model that gives the reliability estimates for flexible pavements. But simulation time is very high which is a hindrance for its practical applications. To encounter this problem, I am planning to obtain fragilities for different design variables and given demand values. For instance, there will many fragility estimates for different thickness values. To obtain a close for this equation, I want to fit a regression model on the obtained fragilities. I want your suggestions about how to proceed and which regression model will be best suited. According to me, as the failure data is in the form of binary variables, I think Logistic regression will be best suited. But I am also confused about COX regression. Please give your suggestions so that I can proceed with my work
Regards
Vighnesh Deshpande
I have developed a black box model that gives the reliability estimates for flexible pavements. But simulation time is very high which is a hindrance for its practical applications. To encounter this problem, I am planning to obtain fragilities for different design variables and given demand values. For instance, there will many fragility estimates for different thickness values. To obtain a close for this equation, I want to fit a regression model on the obtained fragilities. I want your suggestions about how to proceed and which regression model will be best suited. According to me, as the failure data is in the form of binary variables, I think Logistic regression will be best suited. But I am also confused about COX regression. Please give your suggestions so that I can proceed with my work
Regards
Vighnesh Deshpande