A Hybrid Heuristic Dimensionality Reduction Methods for Classifying Malaria Vector Gene Expression Data

Malaria is the world’s leading cause of death, spread by Anopheles mosquitoes. Gene expression is a fundamental level where the effects of unseen vital revealing genes and developmental systems can be evident for detection of distinctions in malaria infections, to recognize the biological processes in human. Ribonucleic acid sequencing offers a large-scale measurable generated profiling transcriptional data results that help a variety of applications such as scientific and clinical condition studies. A fundamental limitation of ribonucleic acid sequencing consists of high dimensional, infrequent and noises, making classification of genes challenging. Several approaches have proposed enhancing the problem of the curse of dimensionality problem, requiring more improvement, yet it is critical to obtain accurate results. In this study, a hybrid dimensionality reduction technique proposes an optimized Genetic algorithm to pick pertinent subset features from the data. Features chosen is passed into principal component analysis and independent component analysis methods grounded on their class variants, to help transform the selected elements into a lower dimension separately. Support vector machine kernel classifiers used the reduced malaria vector dataset to assess the classification performance of the experiment.