Tag Archives: lognormal model

Comparison on performance of the Lognormal, Log logistic and Weibull Distribution on Survival of HIV Patients with Opportunistic Infections in Anambra State, Nigeria (Published)

This paper examined performance of the Lognormal, Log logistic and Weibull distributions on survival of HIV patients with opportunistic infections with the following specific objectives: to fit model for the survival of HIV patients with opportunistic infections for the log-normal distribution, log-logistic distribution and Weibull distribution and to determine the best distribution model for estimating the survival of HIV patients with opportunistic infections. The R-software is employed for data analysis. The secondary data used was fitted to the Weibull model, lognormal model and the log logistic model for selecting the best model for estimating the survival of HIV patients with opportunistic infections in Anambra State from 2010 to 2018. The Akaike information criteria (AIC) and the sum of squares error (SSE) were adopted for assessing the performance of the models. The findings of the study showed that the Weibull model recorded the least value of AIC of 3090.021, the log logistic model was found to be the second with value of AIC = 3178.795 and  the log normal model showed the highest value of 3203.729.Furthermore, the Weibull model equally showed the least SSE of 1333891, following log normal model that showed SSE = 1458851 and the log logistic model showed the highest sum of square error value of 1719035. From the results of the findings, the Weibull model is selected as the best parametric model for estimating the survival of HIV patients with opportunistic infections in Anambra State.

Keywords: log logistic model, lognormal model, opportunistic infections, weibull model

Comparison on performance of the Lognormal, Log logistic and Weibull Distribution on Survival of HIV Patients with Opportunistic Infections in Anambra State, Nigeria (Published)

This paper examined performance of the Lognormal, Log logistic and Weibull distributions on survival of HIV patients with opportunistic infections with the following specific objectives: to fit model for the survival of HIV patients with opportunistic infections for the log-normal distribution, log-logistic distribution and Weibull distribution and to determine the best distribution model for estimating the survival of HIV patients with opportunistic infections. The R-software is employed for data analysis. The secondary data used was fitted to the Weibull model, lognormal model and the log logistic model for selecting the best model for estimating the survival of HIV patients with opportunistic infections in Anambra State from 2010 to 2018. The Akaike information criteria (AIC) and the sum of squares error (SSE) were adopted for assessing the performance of the models. The findings of the study showed that the Weibull model recorded the least value of AIC of 3090.021, the log logistic model was found to be the second with value of AIC = 3178.795 and the log normal model showed the highest value of 3203.729.Furthermore, the Weibull model equally showed the least SSE of 1333891, following log normal model that showed SSE = 1458851 and the log logistic model showed the highest sum of square error value of 1719035. From the results of the findings, the Weibull model is selected as the best parametric model for estimating the survival of HIV patients with opportunistic infections in Anambra State.

Keywords: loglogistic model, lognormal model, opportunistic infections, weibull model