Improving Accurate Candidates for Missing Data Using Benefit Performance of (ML-SOM)

Authors

Abeer Abdullah AL-Mohdar, Faculty of Computers and Information Technology, Hadhramout University.
Mohamed Abdullah Bamatraf, Faculty of Computers and Information Technology, Hadhramout University.

Abstract

Missing data is one of the major challenges in extracting and analyzing knowledge from datasets. The performance of training quality was affected by the appearance of missing data in a dataset. For this reason, there is a need for a quick and reliable method to find possible solutions in order to provide an accurate system. Therefore, the previous studies provided robust ability of Self Organizing Map (SOM) algorithm to deal with the missing values [6, 20]. However, it has a drawback such as an error rate(ERR) in the missing values that increase huge dataset. This study is mainly based on the projection of unsupervised Multilayer SOM (ML-SOM) for missing values. The global methodology presented the combination of advantages of the proposed ML-SOM to obtain a precise method with various missing rates: 5%, 10% and 20%. The experiments were conducted by adopting two types of commonly used data benchmarks (IRIS and BreastCancer) from Weka 3.9 machine learning tool. The new proposed method ML-SOM provides a minimum Error Rate (ERR) and high accuracy (ACC)in small and large datasets compared to other standard classifier types (Bayes-Net, Kmeans and SOM).

Recommended Citation

AL-Mohdar, Abeer Abdullah and Bamatraf, Mohamed Abdullah (2020) "Improving Accurate Candidates for Missing Data Using Benefit Performance of (ML-SOM)," Hadhramout University Journal of Natural & Applied Sciences: Vol. 17: Iss. 1, Article 6.
Available at: https://digitalcommons.aaru.edu.jo/huj_nas/vol17/iss1/6