A Novel Approach to Missing Data Estimation Technique for Microarray Gene Expression Data and Dimensionality Reduction
K Ishthaq Ahmad1, Shaheda Akthar2
1K Ishthaq Ahmad, Research Scholar, Dept. of Computer Science and Engineering, Acharya Nagarjuna University, Guntur (Andhra Pradesh), India.
2Dr.Shaheda Akthar, Registrar FAC Dr. Abdul Haq Urdu University, Kurnool (Andhra Pradesh), India.
Manuscript received on 02 June 2019 | Revised Manuscript received on 07 June 2019 | Manuscript published on 30 June 2019 | PP: 2098-2108 | Volume-8 Issue-8, June 2019 | Retrieval Number: H7535068819/19©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: Microarray gene expression data analysis is one of the finest areas of gene expression analysis, where each gene with its expression value is useful to decide the future analysis of different genes and its characteristics values. Usually, when a data undergoes analysis consisting of missing values and the analysis performed on this data may lead to inconsistent results. We need to recover all these missing values before performing the data analysis, which incurs in the data set. This paper brings out a new method of missing data estimation with the help of clustering technique like DBSCAN for estimating missing values. We also found similar characteristic gene clustering and applied separately to the missing data estimation on these clusters. So, it is a two-step process of missing data estimation, and has an advantage in the context of data reduction dimensionally and smooths application of missing data estimation algorithm. By conducting, an experiment on two microarray data sets, its result, and performance analysis are recorded.
Keywords: Microarray Gene Expression Data, Imputation, PCA, Miss Forest, DBSCAN, RMSE.
Scope of the Article: Expert Approaches