Bulletin of Computational Applied Mathematics (Bull CompAMa)
The Discrete Modified Lindley Distribution
Lishamol Tomy, Veena G, Christophe Chesneau
The data generated in the world of technology is becoming increasingly complex, making it harder to describe all the information behind real data using standard models. This is especially true for discrete data. In light of this, we suggest a novel discrete model, based on a one-parameter discrete distribution. A series of reliability and statistical properties of this distribution are derived. Closed forms can be used to represent these properties. One of the key advantages of the newly developed model is that it may be employed to represent not only positively skewed data sets, but also increasing, decreasing or unimodal failure rate data sets. Subsequently, the unique parameter is estimated using the maximum likelihood approach. Based on this approach and some related criteria, the flexibility of the model in assessing diverse types of data is demonstrated by comparing two different real data sets. In particular, we show that the model has the ability to describe positive data in numerous fields as a supplement to various literature-based models.
Keywords: Discretizing; estimation; modified Lindley distribution; discrete distribution; real data analysis.