An Application of Burrxii-Dal and Weibull-Dal Distributions to Investigates Bank Customers Waiting Time
Background: Data distribution is highly complex and needed a more authentic method to determine by using different distribution models. BURRXII-DAL and WEIBULL-DAL distributions are one if the main models that are used. Objective: The purpose of the study is to determine application of BURRXII-DAL...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Info Sci Publisher
2021
|
| Subjects: | |
| Online Access: | https://repo.uum.edu.my/id/eprint/30947/1/W%2018%2005%202021%202103-2115.pdf https://repo.uum.edu.my/id/eprint/30947/ http://www.webology.org |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Background: Data distribution is highly complex and needed a more authentic method to determine by using different distribution models. BURRXII-DAL and WEIBULL-DAL distributions are one if the main models that are used. Objective: The purpose of the study is to determine application of BURRXII-DAL and WEIBULL-DAL distributions to investigates bank customers waiting time by using different parameters and there comparison with other models.
Methods: BurrXII-DAL and W-DAL distributions each with five parameters are two new lifetime models created from distributions of Burr-G and Weibull-G family. “Burr-XII DAL was compared to six models, which included Kumaraswamy generalised power Weibull (KwGPW, BENH, EGPW, GPW and NH), Beta Exponentiated Nadarajah Haghighi (BENH) and Exponentiated Generalized Power Weibull (EGPW)”. “Weibull -DAL's fit to Weibull Dagum (WDa), Weibull power function (WPF), Weibull Lomax (WLx), generalised power Weibull (GPW), Nadarajah Haghighi (NH), Beta Weibull (BW), and Kumaraswamy Weibull (KwW) distributions”. “One genuine data set is used to test the efficiency of distribution”. “R packages like BFGS (Broyden-Fletcher-Goldfarb-Shanno), SANN (Simulated-Annealing) and NM (Nature-Methods-Methods) are used to process MLEs and the goodness of fit measures containing statistics Criterion such as Akaike Information (AIC), Bayesian Information (BIC), Anderson-Darling (A*), Cramer–von Mises (W*), and K-S statistics (Nelder-Mead)”. Result: “The results of criteria’s including both Akaike information (AIC) and Bayesian information (BIC) of each of the six models showed that the predicted Burr-XII DAL is as best as the other models comparatively”. MLE using Weibull-DAL are small enough with smaller standard errors in variation among original and fitted values of model in presence of these parameters. Conclusion: The parameters generated using MLE and goodness of fit techniques have produced Weibull-DAL as the best distribution among all distributions in term of parameters of model estimating |
|---|