倒傳遞類神經網路處理不平衡大數據效能提升之研究

The study of improving imbalanced big data performance with BPNN approac

鍾吉誠(Chi-Cheng Chung)

指導教授 ： 林真伊

共同指導教授 ： 邱淑芬

國立臺中科技大學/資訊與流通學院/資訊管理系碩士班/碩士(2017年)

摘要

在數據急遽成長的時代，巨量資料已成為目前較常見的議題。巨量資料的分析中，不平衡數據也存在其中。然而在數據不斷的成長情況下，若是我們不修改我們現有的演算法，將會導致我們現在的演算法無法適應更為龐大的數據，也沒有辦法提供較佳的響應時間。所以在計畫中我們預計將倒傳遞類神經網路移植到目前較流行的Spark 運算平台上，減少倒傳遞類神經網路在進行大量數據分類的響應時間，以及透過成本敏感的方法，結合倒傳遞類神經網路提出新的演算法，提升倒傳遞類神經網路進行分類不平衡數據的效能。

關鍵字

巨量資料 ； 不平衡數據 ； 倒傳遞類神經網路

並列摘要

In the data era of rapid growth, big data has become more common issues. Analysis of big data, imbalanced data are also present. However, in the case of data continue to grow, if we do not modify our existing algorithms will lead to our current algorithms can not adapt to big data, there is no way to provide a better response time. So we plan expected to back-propagation neural network migration to the current more popular Spark computing platforms, reducing back-propagation neural network response time performed big data classification, and through cost-sensitive approach, combined with back-propagation neural network to propose new algorithms enhance back-propagation neural network to classify the imbalanced data performance.

並列關鍵字

Big data ； imbalanced big data ； back-propagation neural network

參考文獻

[2] Mayer-Schönberger, V., & Cukier, K. (2013). Big data: A revolution that will transform how we live, work, and think. Houghton Mifflin Harcourt

[3] Ma, Y., Wu, H., Wang, L., Huang, B., Ranjan, R., Zomaya, A., & Jie, W. (2014). Remote sensing big data computing: Challenges and opportunities.Future Generation Computer Systems.

[4] Gupta, R., Gupta, H., & Mohania, M. (2012). Cloud computing and big data analytics: what is new from databases perspective?. In Big Data Analytics (pp. 42-61). Springer Berlin Heidelberg.

[5] Chen, C. P., & Zhang, C. Y. (2014). Data-intensive applications, challenges, techniques and technologies: A survey on Big Data. Information Sciences,275, 314-347.

[6] López, V., Fernández, A., Moreno-Torres, J. G., & Herrera, F. (2012). Analysis of preprocessing vs. cost-sensitive learning for imbalanced classification. Open problems on intrinsic data characteristics. Expert Systems with Applications,39(7), 6585-6608.