Missing Data Estimation And Modelling Via Probability Distributions : A Case Study Of Air Pollutant Data In Malaysia

This book provides simple applications of single imputation methods in replacing missing values and latermodel the dataset using common probability distributions i.e. Weibull, gamma, lognormal etc. The first chapter reviews the theory on missing data mechanism, single and multiple imputations and the types of software available to fill the missing data. Then, the assessment of few single imputation methods were described in chapter 2. The selections of the most appropriate method for the observations were also revealed. In the last chapter, the readers will be exposed on how to model the air pollutant data using probability distributions. The most fitted distributions were also selected after calculating the performance indicators and finally, the most fitted distribution will be used to estimate the return period for next year. Hopefully with this humble publication, the interest among the readers will be developed to explore this new chapter of research thus improving the quality of dataset for better analysis. Furthermore, the good quality data can be model for the future.