WCSE

Machine Learning Based Gene Data Classification Method Research

Abstract— Gene expression data classification aims to automatically assign categories or classes to unseen gene expression data by using the existing history gene expression data. Artificial neural networks are computational models inspired by the structure of biological neural networks. Artificial neural networks have the advantages of self-adapting, self-organizing, and self-learning, and have the ability to possess robustness, parallelism, and generalization. Ensemble learning is one of the major advances in machine learning, and it is a method to employ multiple learners and then combine their predictions to output the final decision. In this paper, we propose to combine the artificial neural networks and ensemble learning technique to do classification for gene expression data. Experimental evaluation of different methods is performed on public gene expression dataset and the results showed the proposed method achieves significant performance improvement.

Index Terms— artificial neural networks, ensemble learning, gene expression data

Lei Shi, Zeqi Xie, Jianfeng Ren, Yueyun Du, Hao Yuan, Qing Zhang
School of Electronics and Information Engineering, Sias International University, CHINA

[Download]

Cite: Lei Shi, Zeqi Xie, Jianfeng Ren, Yueyun Du, Hao Yuan, Qing Zhang, "Machine Learning Based Gene Data Classification Method Research," Proceedings of 2017 the 7th International Workshop on Computer Science and Engineering, pp. 128-132, Beijing, 25-27 June, 2017.