Microarray Gene Expression Data and Performance Analysis of Various Missing Data Imputation Techniques
K Ishthaq Ahmad1, Shaheda Akthar2
1K Ishthaq Ahmad, Research Scholar, Department of Computer Science and Engineering, Acharya Nagarjuna University, Guntur.
2Dr. Shaheda Akthar, Registrar FAC Dr. Abdul Haq Urdu University, Kurnool.
Manuscript received on 11 April 2019 | Revised Manuscript received on 16 May 2019 | Manuscript published on 30 May 2019 | PP: 3126-3130 | Volume-8 Issue-1, May 2019 | Retrieval Number: A1467058119/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Data with missing value is a curse for valuable data, especially in the case of microarray data analysis. Usually, Microarray gene expression data looks like matrix data with a set of genes under various environmental conditions. Microarray gene expression data further undergoes processing in terms of deviations and some other statistical measures. These statistical measures require microarray gene expression data with complete values, but in general condition, data consisting of a certain percentage of missing values. Results which obtain on these missing gene expression data are inconsistent and this result deviates from the original. So that it is necessary to impute the missing values before any estimation is done. In this paper, we have analyzed the performance of various imputation methods based on two real-time microarray datasets.
Index Terms: Missing Data, Microarray Array, Imputation, Predictive Mean, Random-Forest.
Scope of the Article: Measurement & Performance Analysis