St. James Ballroom
Fitting Skewed Data into Asymmetric Distribution (303801)
Akeem Adekunle Akingbade, Yaba College of Technology*Nofiu Idowu Badmus, Yaba College of Technology
Sesan Oluyomi Olufolabo, Yaba College of Technology
Keywords: Asymmetric, kurtosis, Parametric Distribution, skewness, symmetric
The nature of data may guide the choice of which distribution among a set of parametric distributions. The skewness and kurtosis results are useful for this purpose. Meanwhile, the problem of non-zero skewness reveals a lack of symmetry of the empirical distribution. In this study, the descriptive statistics and Exploratory Data Analysis (EDA) was conducted to know the direction and nature of the data; and to develop an appropriate distribution for the data. However, a robust asymmetric distribution is developed due to the nature of the data and some of its statistical properties are derived. The estimation of parameters is obtained through maximum likelihood. The performance of the developed distribution was determined using secondary data on time (months) some patients with tuberculosis were admitted at Lagos University Teaching Hospital (LUTH), Lagos; using a sample of 264 patients. Also, Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC) and Consistent Akaike Information Criterion (CAIC) were used for assessment. Therefore, from all indication, the developed asymmetry distribution is appropriate and capable of handling such data.