Browsing by Author "Odhiambo, Collins"

Now showing 1 - 4 of 4
  • Thumbnail Image
    Item
    A smooth test of goodness-of-fit for the baseline hazard function in recurrent event models
    Odhiambo, John W.; Odhiambo, Collins; Omolo, Bernard
    In this paper, we formulate a smooth test of goodness-of-fit for a simple hypothesis about the baseline hazard function in recurrent-event models. The formulation is an extension of Neyman' s goodness-of-fit approach, whose score tests are obtained by embedding the null hypothesis in a larger class of hazard rate functions. Since the application is in recurrent event models , the data is dynamic.A useful feature about this test is the parametric approach that makes inference about the hazard function more efficient. To examine the finite-sample properties of this test, we used simulated data . For validation, we applied the test to a real-life recurrent event data. Results show that the test possesses better power over wide range of alternatives, when compared with similar tests of the chi-square type in the literature.
  • Thumbnail Image
    Item
    A comparative evaluation of goodness-of-fit tests for the negative binomial distribution with application to RNA-Seq data
    (Strathmore University, 2019-08) Osumba, John; Odhiambo, Collins; Omolo, Bernard
    The negative binomial (NB) distribution is considered the most appropriate distribution for modeling over dispersed count data. In this regard, a number of goodness of-fit (GOF) tests have been applied for the NB, but no systematic evaluation of these tests has been done to determine the most powerful tests. In this study, we perform a comparative evaluation of the GOF tests for the negative binomial distribution, based on their power under suitable alternatives, via simulations. The tests considered here include those based on the empirical distribution functions (EDFs), likelihood functions, Kullback-Leibler discrimination information, Laplace transforms, the generalized smooth tests and a combination of tests with complimentary behavior. For illustration and validation, RNA-Seq data from colorectal cancer are used.
  • Thumbnail Image
    Item
    Data driven longitudinal model with application to HIV differentiated care
    (Strathmore University, 2019-08) Odhiambo, Collins; Weunda, Stephen
    Differentiated care is a new innovative approach for managing HIV/AIDS where ART treatment services are customized by staggering patient’s visits for stable status while reducing unnecessary burdens on the health system. Through provision of differentiated care, the health system can reallocate resources to patients most in need who are failing treatment. The main objective of this study is to develop a data-driven longitudinal model which is applicable to HIV differentiated care. Method: We used routine data of HIV positive patients initiated to ART at the point of care from 4 medical facilities in Nairobi in the year 2018. Since both the GLMM and GEE are extensions of the GLM, we start with a brief overview of GEE then relooked at extensions of GLMM. We specify f (u) and g(p) to be dependent on the type of response Yi . For a binary Yi, we consider f (u) as Bernoulli distribution and g(p) as the logit function, g(u) — log resulting to GLM is the logistic regression. Results show the binary response which was differentiated care category fits well with GLMM. We also found TB-HIV co-infection to be the only significant predictor of differentiated care under both GEE and GLMM.
  • Thumbnail Image
    Item
    Validation of the smooth test of goodness-of-fit for proportional hazards in Cancer survival studies
    (Strathmore University, 2017) Odhiambo, Collins; Odhiambo, John; Omolo, Bernard
    In this study, we validate the smooth test of goodness-of-fit for the proportionality of the hazard function in the two-sample problem in cancer survival studies. The smooth test considered here is an extension of Neyman’s smooth test for proportional hazard functions. Simulations are conducted to compare the performance of the smooth test, the data-driven smooth test, the Kolmogorov-Smirnov proportional hazards test and the global test, in terms of power. Eight real cancer datasets from different settings are assessed for the proportional hazard assumption in the Cox proportional hazard models, for validation. The smooth test performed best and is independent of the number of covariates in the Cox proportional hazard models.