Modelling medical data using the cosine generalized exponential distribution

The volume of data accessible for analysis is expanding rapidly, necessitating the development of new probability distributions to better represent each phenomenon or experiment researched. Applied sciences including medicine, engineering, finance, and actuarial science rely on statistical modeling to evaluate unpredictable lifetime data. Emerging healthcare innovations, such as the integration of big data analytics and artificial intelligence has increased the complexity of health data. The complexity of health data necessitates the use of probability models that are combined with trigonometric functions to represent periodic events existing in the data accurately. This paper aimed to extend the generalized exponential distribution by using trigonometric functions without addition of extra parameters to the baseline model. Numerous characteristics of the extended model are studied and graphical analysis demonstrates that the extended model is a right-skewed probability distribution. Maximum likelihood technique was used and estimated performance was measured by using a Monte Carlo simulation study. The efficiency of the proposed distribution was compared with competing models by using three real dataset from the health sector. The result show that cosine-generalized exponential distribution outperformed other competing models in terms of AIC, BIC, Kolmogorov-Smirnov statistic, Cramer-von Mises statistic, and Anderson-Darling statistic.